System for detection of intermittent stationary objects and method to this end

FIELD: physics, navigation.

SUBSTANCE: set of inventions relates to detection of intermittent stationary objects. System to detect intermittent stationary object nearby moving object comprises image capture device, module to convert the point of observation, module to extract characteristic points, module to compute signal shape data, module to detect data on peaks, module to detect versions of intermittent stationary objects and module to evaluate said objects. System to detect intermittent stationary object nearby moving object comprises image capture step, step to convert the point of observation, step to extract characteristic points, step to compute signal shape data, step to detect data on peaks, step to detect versions of intermittent stationary objects and step to evaluate said objects.

EFFECT: high accuracy of detection.

15 cl, 21 dwg

 

The technical field to which the invention relates

[0001] the Present invention relates to a system for the detection of periodic stationary objects and to a method for the detection of periodic stationary objects.

The level of technology

[0002] Previously proposed a system for detection of objects that creates a differential image from several captured images captured by the camera, and when the shape of the area in which there is a difference in the difference image is changed to the main axis of the direction in which the camera captures the image, estimates that the difference shows a stationary three-dimensional object (see patent document 1).

The list of references

Patent literature

[0003] the Patent document 1. Publication of the patent application (Japan) No. 2007-129560

Disclosure of the invention

The technical problem

[0004] the detection of the objects described in patent document 1, estimates that there is or there is only one stationary three-dimensional object. Therefore, it has difficulty in distinguishing and recognition of stationary three-dimensional objects, periodically present along the roadside, such as power pylons, road signs or telephone poles (hereinafter called the periodic stationary objects),� another three-dimensional object.

[0005] the Present invention is made to solve the above problem. The purpose of the present invention is to provide a system for detecting periodic stationary objects and a method for the detection of periodic stationary objects, which provide high-precision detection of periodic stationary objects.

The solution of the problem

[0006] Aspect of the present invention is a detection system of periodic stationary objects to detect periodic stationary object in the vicinity of a moving object. Detection system of periodic stationary objects includes: a capture device image, mounted on a moving object and allowing the image capturing the surroundings of a moving object; a conversion module viewpoint is arranged to convert at the point of view for the images that are captured by a capture device images to create a picture of the view from the height of bird's flight; the extraction module feature points, configured to extract a characteristic point three-dimensional object from image data in a predetermined area of the image with the bird's eye view for each of a number of subfields included in the pre�artelino specific area; module calculate form data signal, configured to calculate data of the waveform corresponding to the distribution of feature points extracted by the extraction module feature points in a predetermined region of the image with the bird's eye view; a detection module information peaks, is arranged to detect the peaks form data signal; a detection module variants periodic stationary objects, performed with the opportunity to assess whether or not a three-dimensional object having the feature point extracted by the extraction module feature points, a variant of the periodic stationary object, based on equals or exceeds either no information peaks predetermined first threshold value; and an evaluation module of the periodic stationary objects, is arranged to determine that a variant of the periodic stationary object is a periodic stationary object when the periodic variant of the stationary object detected by the detection module variants periodic stationary objects, and the detection is carried out when the predetermined condition.

[0007] Another aspect of the present invention is a method detected�tion of periodic stationary objects to detect periodic stationary objects in the vicinity of a moving object. Method of detection periodic stationary object includes: a step of capturing images to capture images of the surroundings of a moving object using the device for capturing images mounted on a moving object; a phase conversion angle conversions point of view for the images that are captured by a capture device images to create a picture of the view from the height of bird's flight; a step of extracting feature points to extract a characteristic point three-dimensional object from image data in a predetermined area of the image with the bird's eye view for each of a number of subfields included in the predetermined region; the step of calculating the data of the waveform to calculate the data waveform corresponding to the distribution of feature points extracted at the stage of extraction of the feature points in a predetermined region of the image with the bird's eye view; a step of detecting information of the peak detection information peaks form data signal; a phase detection options periodic stationary objects to assess whether or not a three-dimensional object having the feature point extracted at the stage of extraction of the characteristic points, variant PHE�ionicheskogo stationary object, on the basis that equals or exceeds either no information peaks predetermined first threshold value; and the evaluation phase of the periodic stationary objects to determine that the periodic stationary object is a periodic stationary object when the option of periodic stationary object is detected in the detection step variants of the periodic stationary objects, and the detection is carried out when the predetermined condition.

Brief description of the drawings

[0008] Fig.1 is a schematic diagram of a configuration of the detection system of periodic stationary objects according to the first embodiment of the present invention, illustrating an example in which the detection system of periodic stationary objects mounted on the vehicle.

Fig.2 is a top view to illustrate the state of motion of the vehicle shown in Fig.1.

Fig.3 is a flowchart for illustrating details of the calculation module shown in Fig.1.

Fig.4 shows the top views to illustrate the General idea of processing by the combining module, shown in Fig.3 (a) illustrates the state of motion of the vehicle and (b) illustrates the General� representation of alignment.

Fig.5 shows views for illustrating details of processing by the module of calculation of options values of displacement, shown in Fig.3 (a) shows the difference image PDtat time t and (b) shows the difference image PDt-1at time t-1.

Fig.6 is a block diagram of the sequence of operations of a method for illustrating processings by means of module combination and module detection of three-dimensional objects shown in Fig.3.

Fig.7 is a block diagram of the sequence of operations of a method for illustrating the processing by the module determine the frequency, shown in Fig.3.

Fig.8 is a graph showing the histogram generated by the counting module shown in Fig.3.

Fig.9 is a block diagram of the sequence of operations of a method for illustrating processings via the automatic calculation of range of motion and assessment module of the periodic stationary objects, shown in Fig.3.

Fig.10 shows drawings for illustrating the details of step S27 shown in Fig.9 (a) illustrates the case in which another vehicle appears in front of periodic stationary objects, (b) shows the histogram in (a), (c) illustrates the case in which another vehicle appears at the opposite one hundred�Onet from periodic stationary objects, and (d) shows a histogram in the case of (c).

Fig.11 is a diagram of a schematic configuration of the detection system of periodic stationary objects according to a second embodiment of the present invention, illustrating an example in which the detection system of periodic stationary objects mounted on the vehicle.

Fig.12 is a view to illustrate the state of motion of the vehicle shown in Fig.11, and the range of capture devices capture images.

Fig.13 is a flowchart for illustrating details of the calculation module shown in Fig.11.

Fig.14 is a view for illustrating detailed operations of the module calculate the distribution of the edges, the counting module, the detection module variants periodic stationary objects and module evaluation periodic stationary objects, shown in Fig.13.

Fig.15 is a block diagram of the sequence of operations of way to illustrate details of the method of detection periodic stationary objects according to a second embodiment of the present invention.

Fig.16 is a block diagram of the sequence of operations of way to illustrate details of the method of detection periodic stationary objects according to a second embodiment of the present invention, illustrierte after Fig.15.

Fig.17 is a flowchart for illustrating details of the module computing system for the detection of periodic stationary objects according to a third embodiment of the present invention.

Fig.18 shows graphs for illustrating detailed operations of the combining module of Fig.17.

Fig.19 shows graphs for illustrating detailed operations of a module for calculating the difference between Fig.17, (a) illustrates the difference in the case that the waveform of the distribution of edges is extracted from the periodic stationary objects, and (b) illustrates the difference in the case that the waveform of the distribution of edges is extracted from moving objects.

Fig.20 is a block diagram of the sequence of operations of way to illustrate details of the method of detection periodic stationary objects according to a third embodiment of the present invention, corresponding to Fig.16.

Fig.21 is a block diagram of the sequence of operations of way to illustrate details of the method of detection periodic stationary objects according to the modified example of the third embodiment of the present invention, corresponding to Fig.16.

The implementation of the invention

[0009] the First variant implementation

Further in this document describes the preferred embodiments carried�of tvline of the present invention based on the drawings. Fig.1 is a schematic diagram of the system configuration of the 1 detection of periodic stationary objects according to the first embodiment of the present invention, illustrating an example in which the detection system 1 of the periodic stationary objects mounted on the vehicle V. the System 1 detection of periodic stationary objects, shown in Fig.1, is arranged to detect periodic stationary objects in the surroundings of the vehicle V and, in particular, to detect stationary objects, periodically present along the roadside, such as power pylons, road signs or telephone poles. In this regard, the following examples describe the considered vehicle V as an example of a moving object. However, a moving object under consideration is not limited by the vehicle V and may be any other moving object, such as a motorcycle or Bicycle.

[0010] the detection System 1 of the periodic stationary objects includes a chamber 10 (the device image capturing), the sensor 20 vehicle speed (speed detector) and the module 30 of the calculation. The camera 10 shown in Fig.1, is mounted in position at a height h and at the rear of Russ�travelago vehicle V thus, that its optical axis forms an angle θ of inclination downwards relative to the horizontal. The camera 10 is arranged to capture an image of a pre-defined scope of this position. The sensor 20 of the speed of the vehicle is arranged to determine the speed of the vehicle V and calculate the speed from the speed of rotation of the wheels, by means of the gauge of speed of rotation of the wheels provided on the wheel to determine, for example, the number of revolutions. Module 30 results of the calculations with the ability to detect periodic stationary object in the vicinity of the vehicle V on the basis of the image captured by the camera 10, and the signal from the sensor 20, the speed of the vehicle.

[0011] Fig.2 is a top view to illustrate the state of motion of the vehicle V shown in Fig.1. As shown in Fig.2, the camera 10 captures an image of area behind the vehicle under a predefined angle a of the review. In this case, the camera 10 has a wide angle of the review and allow image capture of the lane, which considered moving the vehicle V, as well as by the side of which there is a periodic stationary about�project.

[0012] Fig.3 is a flowchart for illustrating details of the module 30 of the calculation shown in Fig.1. It should be noted that in Fig.3 also illustrates the camera 10 and the sensor 20, the vehicle speed, to clarify the relationship connections.

[0013] As shown in Fig.3, the module 30 of calculations includes a module 31 conversion point of view, the combining module 32 module 33 detection of three-dimensional objects, the module 34 calculation of options values of displacement, the counting module 35 module 36 calculation of range of motion, the module 37 assessment of periodic stationary objects, and the module 38 to the detection of the lane change (detection module transverse motion).

[0014] the Module 31 conversion of a viewpoint arranged to receive the captured image data obtained by capturing by the camera 10, and to convert the viewpoint taken to convert captured image data into image data of a kind from height of the bird's flight in a state of viewing from a bird's eye view. Review status with the bird's eye view means a view from the virtual camera, which is directed, for example, vertically down from the sky. Such a transformation of the viewpoint is performed as described, for example, in patent� document 1.

[0015] the combining Module 32 configured to sequentially receive data of images of views from the bird's eye view obtained through the conversion of the viewpoint by means of module 31 conversion point of view, and combine the position data received images of the view from the bird's eye view at different points in time. Fig.4 shows the top views for illustrating the overall presentation of the processing module 32 through the combination shown in Fig.3 (a) illustrates the state of motion of the vehicle V and (b) illustrates an overall view of alignment.

[0016] As shown in Fig.4(a), suppose that the vehicle V at the current time is in V1and was in V2during one time segment before. In addition, other vehicle VOis in the rear side of the vehicle V and consider moving parallel to the vehicle V provided that the other vehicle VOat the current time is in VO1and was in VO2during one time segment before. Additionally, it is acceptable that the vehicle V moves a distance d in a time segment. It should be noted that the time at one time �egment earlier may be the time for a predetermined time period (for example, one control cycle) earlier than the current time or may be the time for an arbitrary period of time before the current time.

[0017] In this state, the image PBtkind from height of bird's flight at the current time is as shown in Fig.4(b). In the image PBtkind from height of bird's flight white line road markings painted on the road, is rectangular in shape and is relatively accurate view from the top. Meanwhile, another vehicle VOlocated in the VO1seen at an angle. In addition, also in the image PBt-1kind from height of bird's flight during one time segment before the white line road markings painted on the road, is rectangular in shape and is relatively accurate view from the top. However, the other vehicle VOlocated in the VO2seen at an angle.

[0018] the combining Module 32 combines image PBt, PBt-1kind from height of bird's flight, as described above, in the data. In this case, the combining module 32 shifts the image PBt-1kind from height of bird's flight during one time segment used to align its position with the position of the image PBtview bird's eyeat the current time. The value of d' �is substituted by a value the corresponding travelled distance d, shown in Fig.4(a), and determined on the basis of the signal from the sensor 20, the vehicle speed and time period with time for a single time segment before the current time.

[0019] the Module 33 to the detection of three-dimensional objects made with the possibility to detect several three-dimensional objects from data on the difference image PDt. In particular, the module 33 to the detection of three-dimensional objects explains the difference between images of PBt, PBt-1kind from height of the bird's flight and creates the data on the difference image PDt. Here, PixelNet value for the difference image PDtcan be set by finding the absolute value of the difference between pixelenemy values of images of PBt, PBt-1kind from height of bird's flight, or can be set as "1" when the absolute value exceeds a predefined value, and as "0" if does not exceed a predefined value, to adapt to the change of the environment light. In addition, the module 33 to the detection of three-dimensional objects made with the ability to evaluate what three-dimensional objects present in the area detected as "1" in the above manner, in the data on the difference image PDt.

[0020] it is about�utitsa again to Fig.3. Module 34 calculation of options values move made with the possibility of options to calculate the magnitude of displacement of several three-dimensional objects detected by the module 33 to the detection of three-dimensional objects. Fig.5 shows views for illustrating details of processing by the module 34 calculation of options values of displacement, shown in Fig.3 (a) shows the difference image PDtat time t and (b) shows the difference image PDt-1at time t-1.

[0021] first, the module 34 to calculate the options value detects movement of the reference point of ground contact (characteristic point) of the three-dimensional object from the difference image PDt-1at time t-1, as shown in Fig.5(b). Datum point ground contact means the contact point between the three-dimensional object and the road. In this case, the module 34 calculation of options values move detects, as the reference points of contact with the ground, nearest the position of the detected three-dimensional object to the camera 10 of the vehicle V. the Module 34 to calculate the options value detects movement of the reference point of ground contact in each region (sub-regions) having a three-dimensional object present in the data on the difference image PDt-1that is estimated at portstatus 33 detection of three-dimensional objects.

[0022] In particular, the module 34 to calculate the options value detects movement of the reference point P1ground contact for a three-dimensional object O1detects the reference point P2ground contact for a three-dimensional object O2and detects the reference point P3ground contact for a three-dimensional object O3. Then the module 34 calculation of options values move sets a region T having a width W in the difference image PDtat time t, as shown in Fig.5(a). In this case, the module 34 of computation variants of displacement magnitude specifies the area of T in positions corresponding to reference points P1-P3contact with the ground, in the data on the difference image PDt-1at time t-1.

[0023] Then, the module 34 to calculate the options value detects movement of the reference point of ground contact of the three-dimensional object from the difference image PDtat time t. In this case, the module 34 to calculate the options value detects movement of the reference point of ground contact in each region (sub-regions) having a three-dimensional object present in the data on the difference image PDtthat module is assessed by means of 33 detection of three-dimensional objects. Module 34 calculation of options values detects movement near PT�th detected three-dimensional object, as a datum point of contact with the ground, to the camera 10 of the vehicle V. In particular, the module 34 to calculate the options value detects movement of the reference point P4ground contact for a three-dimensional object O4,detects the reference point P5ground contact for a three-dimensional object O5and detects the reference point P6ground contact for a three-dimensional object O6. Thus, the combining module 32 module 33 detection of three-dimensional objects and the module 34 calculation of options values of displacement are characteristic extraction module dots configured to extract a characteristic point (datum point of contact with the ground) three-dimensional object from image data in a predetermined area of the image with the bird's eye view (image data in the rear lateral region of the difference image) for each of the multiple sub-regions (each region having estimated the three-dimensional object present in the image data in the difference image) included in the predetermined region.

[0024] Additionally, the module 34 calculation of options values move connects the reference points of ground contact with each other. In particular, the module 34 to calculate the options value�us move connects the reference point P 4ground contact with the reference point P1ground contact that connects the reference point P5ground contact with the reference point P1contact with the land and binds the reference point P6ground contact with the reference point P1contact with the ground. Similarly, the module 34 calculation of options values move connects the reference points P4-P6ground contact with the control points P2and P3contact with the ground.

[0025] After that, the module 34 calculation of options values calculates the moving distance (i.e., variants of displacement magnitude) between reference points P1-P6ground contact connected in this way. Then the module 34 calculation of options values move sets the calculated distance as options of displacement magnitude. Thus, the module 34 calculation of options values move calculates several variants of displacement magnitude for each three-dimensional object. This prevents the problem of an erroneous calculation of displacement magnitude of the periodic stationary object with similar features that occur periodically as a result of the exceptional determination of the three-dimensional movement of the object.

[0026] it Should be noted that the reason that sets the scope of the T, is that sales� if an error occurs with a combination of images PB t, PBt-1kind from height of bird's flight due to the "galloping", deviations in the vertical axis, etc. of the vehicle V, the reference points P1-P6ground contact is stably associated with each other. Additionally, binding of the reference points P1-P6contact with the ground is determined by the reconciliation process for the brightness distribution around the reference points of contact with the ground images of PBt, PBt-1kind from height of bird's flight.

[0027] it is important to refer again to Fig.3. Counter module 35 is arranged to calculate the options value of the displacement calculated by the module 34 of computation variants of displacement magnitude, and using the count of the histogram is formed (data waveform). For example, the audit module 35 calculates the value as "3" when the distance between the reference point P1ground contact and reference point P4contact with the ground, the distance between the reference point P2ground contact and reference point P5ground contact and the distance between the reference point P3ground contact and reference point P6ground contact are identical. Thus, by counting the variants of displacement magnitude and the formation of the histogram, calculating module 35 serves as a module �of ycycline data waveform, configured to calculate data waveform (based on the relative mutual position of the reference points of contact with the ground), corresponding to the distribution of the datum points of contact with the ground in the rear lateral region of the difference image.

[0028] the Module 36 for calculating the range of movement is arranged to calculate the range of motion of the periodic stationary object for the image view from the height of bird flight based on the interval of capture of the camera 10 and the speed of the vehicle V is determined by means of the sensor 20 of the vehicle speed. More specifically, the module 36 for calculating the movement range calculates the range of motion with the correct stock is in a predefined range for the speed of the vehicle V. Here, the allowable margin is, for example, ±10 km/h. More specifically, the module 36 for calculating a range of motion calculates the speed of a three-dimensional object, moving one pixel in one cycle management at about 5.5 km/h, when the interval of capture of the camera 10 is 33 MS, and the actual distance in the direction of movement of the vehicle that are covered by one pixel, is 5 cm, the accuracy of images of PBt, PBt-1kind from height of the bird's flight is reduced due to the movement of a vehicle requires a valid margin of ±10 km/h to make these valid approximately 5.5 km/h.

[0029] the Module 37 assessment of periodic stationary objects is performed with the opportunity to assess whether or not a multiple of three-dimensional objects detected by the module 33 to the detection of three-dimensional objects, the periodic stationary objects. Module 37 assessment of periodic stationary objects includes a module 37a detect variants of periodic stationary objects, and the module 37b determine the frequency. Module 37 assessment of periodic stationary objects is performed with the opportunity to assess whether or not a multiple of three-dimensional objects detected by the module 33 to the detection of three-dimensional objects, the periodic stationary objects, on the basis of the histogram generated by the calculating module 35, range of movement calculated by the calculation module 36 range of movement options periodic stationary objects stationary objects, which can be periodic stationary objects) detected by the module 37a detect variants of periodic stationary objects, and periodicity defined by m�module 37b determine the frequency.

[0030] the following describes a method for the detection of periodic stationary objects in relation to the block diagrams of the sequence of operations of way. Fig.6 is a block diagram of the sequence of operations of a method for illustrating the processing by the combining module 32 and module 33 detection of three-dimensional objects shown in Fig.3. First, the combining module 32 receives image data PBt, PBt-1kind from height of bird's flight at various time points defined by the module 31 conversion point of view, to combine (S1). Then the module 33 to the detection of three-dimensional objects explains the difference between the data image PBt, PBt-1kind from height of bird's flight, combined in step S1 (S2). After that, the module 33 to the detection of three-dimensional objects performs the conversion to binary form based on a predetermined value and generates the data on the difference image PDt(S3). Thus, the end of processing by the combining module 32 and module 33 detection of three-dimensional objects.

[0031] Fig.7 is a block diagram of the sequence of operations of a method for illustrating the processing by the module 37a detect variants of periodic stationary objects and module 37b determine the frequency, shown in Fig.3. Fig.8 is a graph�m to illustrate the histogram, formed through the calculating module 35 shown in Fig.3. As shown in Fig.8, the counting module 35 calculates identical versions of displacement magnitude computed for them. In particular, since several of the variables m1, m2, m3, m4 displacement detected in the example shown in Fig.8, these counter values are high.

[0032] As shown in Fig.7 and 8, the module 37a detect variants of periodic stationary objects first detects the maximum value of M (the peak value, peak) of the histogram (S11). Then the module 37a detect variants of periodic stationary objects specifies a pre-defined threshold value Th1on the basis of the maximum value M detected in step S11 (S12). Here, the predetermined threshold value Th1is set equal to 70% of the maximum value M. for Example, when the value of the counter maximum value M is "7", pre-defined threshold value Th1is set to "4,9". Because the pre-defined threshold value Th1is obtained from the maximum value M of the counter values so you can set a proper threshold value, even if the size of the counter values change significantly due to the mutual arrangement between rassmotrev�seen the vehicle V and three-dimensional objects, solar lighting, etc. it Should be noted that in this embodiment of the pre-defined threshold value Th1is the value in 70% of the maximum value M, but is not limited to this.

[0033] Then the module 37a detect variants of periodic stationary objects detects a local maximum value M1-M3 (peak; peaks of equal to or greater than a predetermined threshold value Th1(S13). Here, when the maximum value M is, for example, "7", module 37a detect variants of periodic stationary objects detects a local maximum value M1-M3, with a value at "5" or more. Thus, the module 37a detect variants of periodic stationary objects acts as a detection module information peaks, is arranged to detect the peaks of the histogram (waveform). Additionally, the module 37a detect variants of periodic stationary objects assesses whether or not a three-dimensional object having the detected fiducial point of contact with the ground, a variant of the periodic stationary object, on the basis that equals or exceeds either no information peaks predetermined threshold value. In castnet�, for example, the module 37a detect variants of periodic stationary objects determines what three-dimensional objects associated with variations of displacement magnitude corresponding to the local maximum values M and M1-M3 (which includes the maximum value M) (for example, when the distance defined between two fixed points in contact with the ground matches any of the local maximum values M and M1-M3, three-dimensional objects represent two of the three-dimensional object having a defined point of contact with the ground), are variants of periodic stationary objects.

[0034] thereafter, the module 37b determine the frequency of detects intervals (peaks) of the local maximum values M and M1-M3 (which includes the maximum value M), and casts the vote on the detected intervals (S14). In particular, in the example shown in Fig.8, the number of votes for the interval D1is "2" and the number of votes for the interval D2is "1".

[0035] Then the module 37b determine the frequency determines whether or not the frequency (S15). In this case, the module 37b determine the frequency determines the frequency based on equals or exceeds or not the number of votes in step S14 pre-determined number of votes. Here, a predetermined number of votes �leaves half of the number of three-dimensional objects, detected from the image PBtkind from height of bird's flight. Therefore, when the number of three-dimensional objects detected from an image PBtkind from height of the bird's flight is "4", pre-determined number of votes is "2". It should be noted that a predetermined number of votes is not limited to the above and may be a fixed value.

[0036] When evaluated that there is a periodicity (S15: Yes), the module 37b determine the frequency lowers the predetermined threshold value Th1in step S12 (S16). Then, the processing proceeds to step S17. Therefore, although the pre-defined threshold value Th1is, for example, 70% of the maximum value M, the pre-defined threshold value Th1is set equal to 60% of the maximum value M or other values. In addition, the period during which the reduced pre-defined threshold value Th1approximately in the region of 1 second. Every time appreciated that there is a periodicity, reset the pre-defined threshold value Th1. Thus, whether or not the frequency is estimated from the positions in which there are local maximum values M and M1-M3 values from water�and, i.e. intervals. When evaluated that there is a periodicity decreases a predetermined threshold value Th1. Accordingly, once determined by the frequency estimation of periodic stationary objects can be simplified. Meanwhile, until not specified the frequency goes below a predetermined threshold value Th1. This prevents erroneous determination of three-dimensional objects due to alignment errors, etc.

[0037] meanwhile, when it is estimated that there is no periodicity (S15: No), the processing proceeds to step S17 without decreasing the predetermined threshold value Th1.

[0038] Thus, the module 37b determine the frequency estimates that a periodicity of the number of votes (information peaks) on the positions in which there are local maximum values M and M1-M3 (intervals), and the local maximum values M and M1-M3 are equal to or exceed a predetermined threshold value Th1on the basis of the maximum value M of the counter values of the options of displacement magnitude. This allows you to ignore the local maximum value, which is a relatively small value (for example, the M4 with reference number in Fig.8), and the frequency can be estimated even t�knee with little influence of noise.

[0039] In step S17, the module 37b determine the frequency estimates, detects or not the module 38 to the detection of the lane change lateral movement by a specified distance or more (S17). In particular, the module 38 to the detection of the lane change estimates that the transverse motion is detected by a specified distance or more when the turn signal, and the detected steering angle is equal to or greater than the specified angle, determined from the vehicle speed defined by the speed sensor of the vehicle.

[0040] In the case of measuring what was detected lateral movement by a specified distance or more (S17: Yes), when a pre-defined threshold value Th1lowered in step S16, the module 37b determine the frequency of low initializes the threshold value Th1(S18). This allows detection of periodic stationary objects appropriately in accordance with change of environment after being replaced lane. After this is completed, the processing shown in Fig.7. Meanwhile, when evaluated that is not detected lateral movement by a specified distance or more (S17: No), the processing shown in Fig.7, initialization completes without a pre-defined threshold Th1.

p> [0041] Fig.9 is a block diagram of the sequence of operations of a method for illustrating the processing by the module 36 for calculating a range of motion and module 37 assessment of periodic stationary objects, shown in Fig.3. As shown in Fig.9, first, the module 36 for calculating a range of motion calculates the amount of movement corresponding to stationary objects (S21). In particular, the module 36 for calculating the movement range calculates the range of motion of the periodic stationary object for the image view from the height of bird flight based on the interval of capture of the camera 10 and the speed of the vehicle V is determined by means of the sensor 20 of the vehicle speed. In this case, the module 36 for calculating the movement range calculates the range of motion with the correct stock is in a predefined range for the speed of the vehicle V.

[0042] Then, the module 37 assessment of periodic stationary objects determines what version of the periodic stationary object is a periodic stationary object when the option of periodic stationary object is detected by the module 37a detect variants of periodic stationary objects, and detection is carried out with preliminary�flax certain condition. In particular, the module 37 assessment of periodic stationary objects evaluated the present or do not have a local maximum values M and M1-M3 (peaks in the histogram) in the range of magnitude of displacement calculated in step S21 (range of movement) (S22). When determining that any of the local maximum values M and M1-M3 is present in the range of displacement magnitude (S22: Yes), the module 37 assessment of periodic stationary objects assesses that the periodic stationary object is present (i.e., determines what version of the periodic stationary object detected by the module 37a detect variants of periodic stationary objects, is a periodic stationary object) (S23). In other words, the periodic stationary objects are often combined through the identical timeframe, and specific counter value tends to be large. In addition, since the periodic stationary object is stationary, the value of the counter options the magnitude of the movement should be in the range of motion specified with the speed of a moving object, etc. Thus, when in step S22 is determined "Yes", we can say that several three-dimensional objects are periodic stationary objects. After this is completed, the processing shown in Fig.9.

[043] meanwhile, when determining that any of the local maximum values M and M1-M3 is not present in the above range of displacement magnitude (S22: No), the module 37 assessment of periodic stationary objects evaluates it, determines whether or not the module 37b determine the frequency that there is a periodicity (S24). When determining that the module 37b determine the frequency does not determine that there is a periodicity (S24: No), the module 37 assessment of periodic stationary objects estimates that a three-dimensional object is a moving object (S25). After this is completed, the processing shown in Fig.9.

[0044] When determining that the module 37b determine the periodicity determines that there is a periodicity (S24: Yes), the module 37 assessment of periodic stationary objects detects aperiodic local maximum value of the local maximum values equal to or greater than a predetermined threshold value Th1(S26). Aperiodic local maximum value is, for example, the local maximum value of M3 shown in Fig.8. The interval of this local maximum value M3 to adjacent local maximum values is different from the spacing for other local maximum values M, M1, M2. Therefore, the module 37 periodic evaluation staz�anenih object determines the what is a local maximum value M3 is aperiodic local maximum value, non-periodicity.

[0045] Additionally, when not detected aperiodic local maximum value (S26: No), there is the frequency, and is not present aperiodic local maximum value. Accordingly, the module 37 assessment of periodic stationary objects assesses that the periodic stationary object is present (S23).

[0046] meanwhile, when the detected aperiodic local maximum value (S26: Yes), the module 37 assessment of periodic stationary objects evaluates, below, or no periodic local maximum values M, M1, M2 previous values (S27). In this arrangement, the module 37 assessment of periodic stationary objects calculates the average value of the periodic local maximum values M, M1, M2 in the current processing, and calculates the average value of the periodic local maximum values in the previous processing. Then the module 37 assessment of periodic stationary objects evaluates, below or not the average value in the current processing of the average value in the previous processing to a predetermined value or more.

[0047] When determining that the periodic local maximum values, M1, M2 lower than previous values (S27: Yes), the module 37 assessment of periodic stationary objects assesses that another vehicle, etc. appears between the vehicle V and the periodic stationary objects, and detects a moving object (S25). After this is completed, the processing shown in Fig.9.

[0048] meanwhile, when determining that the periodic local maximum values M, M1, M2 is not lower than previous values (S27: No), the module 37 assessment of periodic stationary objects assesses that another vehicle, etc. appears on the opposite side of the periodic stationary object when viewed from the side of the vehicle V, and detects the periodic stationary object (S23). After this is completed, the processing shown in Fig.9.

[0049] Fig.10 shows drawings for illustrating the details of step S27 shown in Fig.9 (a) illustrates the case in which the other vehicle VOappears in front of periodic stationary objects; (b) shows the histogram in (a). In addition, (c) illustrates the case in which the other vehicle VOappears on the opposite side of the periodic stationary objects, and (d) shows a histogram in the case of (c). It follows from�etit, what in Fig.10(b) and (d), the dotted line represents the histogram before you receive another vehicle, and the solid line represents the histogram after you receive another vehicle.

[0050] first, assume that another vehicle VOappears in front of periodic stationary objects, as shown in Fig.10(a). In this case, since the periodic stationary objects are locked by another vehicle VOthe count value of the periodic local maximum value tends to become smaller, as shown in Fig.10(b). In particular, when the other vehicle VOappears in front of periodic stationary objects, other vehicle VOmight be in the position to which a given vehicle V can change the lane. Therefore, in this case, the module 37 assessment of periodic stationary objects detects another vehicle VO(a moving object).

[0051] in contrast, assume that another vehicle VOappears on the opposite side of the periodic stationary objects, as shown in Fig.10(c). In this case, the periodic stationary objects are not locked by another tra�sports V O. Thus, the impact on the value of the counter of the periodic local maximum values are practically non-existent, and the count value of the periodic local maximum value becomes not too small. When the other vehicle VOappears on the opposite side of the periodic stationary objects, other vehicle VOis not in the position to which a given vehicle V can change the lane. Accordingly, in this case, the module 37 assessment of periodic stationary objects detects periodic stationary objects.

[0052] In the system 1 detection of periodic stationary objects and the detection of periodic stationary objects in accordance with this embodiment of the reference point of ground contact (characteristic point) of the three-dimensional object is extracted from image data in the difference image in the rear lateral region (predetermined region) of the image with the bird's eye view for each area, in accordance with the definition of a three-dimensional object present in the image data in the difference image for each of a number of subfields included in the predetermined region); calculates the histogram (Dan�s waveform), corresponding to the distribution of the datum points of contact with the ground in rear of the transverse region for the image kind from height of bird's flight; and that, whether or not the three-dimensional object having the extracted reference point of ground contact, the option of periodic stationary object is determined on the basis that equals or exceeds either no information peaks (peak value, the number of votes intervals of peaks, etc.) of the histogram of the pre-defined threshold value. Therefore, the detection system 1 of the periodic stationary objects and a method for the detection of periodic stationary objects provide a more precise extraction of the periodicity (frequency) of periodic stationary objects as information peaks form data signal, and variants of the periodic stationary objects can easier be extracted from three-dimensional objects included in the captured image. Therefore, accurate extraction of periodic stationary objects.

[0053] the Periodic stationary objects are often stationary objects that have similar appearances and a combined through almost equal intervals. When the capture device captures images of the image such periodic stationary objects when driving, it is difficult tale�th what are the elements of the periodic stationary objects in the previous image correspond to which fragments in the current image. In addition, in this case also it is difficult to determine that such captured periodic stationary objects are stationary objects or moving objects. Additionally, the periodic stationary objects may be erroneously recognized as moving objects, depending on conditions such as the speed of movement of a moving object, the interval of capture devices capture images and the deviation on the horizontal axis of the periodic stationary objects.

The detection system 1 of the periodic stationary objects and a method for the detection of periodic stationary objects in accordance with this embodiment of the provide a more accurate extraction of periodic stationary objects from three-dimensional objects included in the captured image, and allow the erroneous detection of periodic stationary objects as moving objects, as described above.

[0054] the Periodic stationary objects form a differential area periodically present in the differential image. It is also difficult to calculate the magnitude of the move when the differential binding of these periodic regions properly with resp�corresponding fragments in the previous image, and it is difficult to determine, exist or not stationary objects.

In the detection system 1 of the periodic stationary objects and the detection of periodic stationary objects in accordance with this embodiment of the options are calculated magnitude of displacement detected several three-dimensional objects are calculated and computed versions of displacement magnitude. Accordingly, the counting is performed when it is unclear what the periodic region of the difference properly correspond in fragments in the previous image. Then, when determined that the counter value in the range of movement of a moving object from the counted values of the counter options the magnitude of displacement is equal to or exceeds the threshold value Th1determined that several three-dimensional objects are periodic stationary objects. Here, the periodic stationary objects are often combined through the identical timeframe, and specific counter value tends to be large. In addition, because of the periodic stationary objects are stationary, the counter values of the options the magnitude of the movement should be in the range of motion specified with the speed of a moving object, etc. Thus, when a particular counter value in the range�the zone of displacement, a predetermined speed of a moving object, etc., is equal to or exceeds a predetermined threshold value Th1you can say that a few three-dimensional objects are periodic stationary objects. Therefore, it is possible to determine with higher accuracy the periodic stationary objects.

[0055] Optionally, the detection system 1 of the periodic stationary objects and the detection of periodic stationary objects according to this variant implementation, several variants of translation amounts are calculated for each three-dimensional object. This allows to prevent erroneous calculations of displacement magnitude of the periodic stationary objects with similar features that occur periodically as a result of an exceptional determine the magnitude of the displacement of three-dimensional objects.

[0056] in Addition, in the system 1 detection of periodic stationary objects and the detection of periodic stationary objects in accordance with this embodiment of the pre-defined threshold value Th1is obtained from the maximum value M of these counted values of the counter. Accordingly, it is possible to set a threshold Th1properly, even if RA�measures of counter values change significantly due to the mutual arrangement between the moving object and three-dimensional objects, solar lighting, etc.

[0057] in addition, the detection system 1 of the periodic stationary objects and the detection of periodic stationary objects according to this variant of implementation, whether or not the frequency is determined from the positions in which there are local maximum values M and M1-M3 from the counted values of the counter. When determined that there is a periodicity decreases a predetermined threshold value Th1. Accordingly, once determined by the frequency determination of periodic stationary objects can be simplified. Meanwhile, until not specified the frequency goes below a predetermined threshold value Th1. This prevents erroneous determination of three-dimensional objects due to alignment errors, etc.

[0058] in addition, the detection system 1 of the periodic stationary objects and the detection of periodic stationary objects according to this variant implementation, the frequency is determined from the positions in which there are local maximum values M and M1-M3, and the local maximum values M and M1-M3 are equal to or exceed a predetermined threshold value Th1on the basis of a four-s� value M of the counter values. This allows you to ignore the local maximum value, which is a relatively small value, and the frequency can be estimated more precisely with a small influence of noise.

[0059] in addition, the detection system 1 of the periodic stationary objects and the detection of periodic stationary objects according to this variant implementation, when the detected lateral movement by a specified distance or more, and lowered the pre-defined threshold value Th1that is initialized to a low threshold value Th1. Thus, the threshold value Th1is initialized when the vehicle V is replaced by the lane. This allows detection of periodic stationary objects appropriately in accordance with change of environment once gives way to the lane.

[0060] Optionally, the detection system 1 of the periodic stationary objects and the detection of periodic stationary objects according to this variant of implementation, in the case of determining a local maximum value of M3 equal to or above the predetermined threshold value Th1than the local maximum values M, M1, M2, having according to the definition of the correct frequency processing, when the average value of the local maximum values M, M1, M2, having according to the definition of the periodicity in the current processing, not less than the average of the local maximum values, in accordance with the definition of frequency in the previous processing, to a predetermined value or more, determines that the multiple three-dimensional objects are periodic stationary objects. Meanwhile, when the first is less than the second pre-defined value or more, determines that the moving object is present.

[0061] moreover, here, as a case to determine the local maximum value of M3 equal to or above the predetermined threshold value Th1than the local maximum values M, M1, M2, having according to the definition of the periodicity in the current processing, a case is possible when, for example, another vehicle, etc. appears within the viewing angle. Such a case may includes a case in which another vehicle, etc. appears on the opposite side of the periodic stationary objects when viewed from the side of the vehicle V, and a case in which another vehicle, etc. appears on the front side.

[0062] When the other transp�comfortable tool, etc. appears on the opposite side, the periodic stationary objects practically does not affect the periodic local maximum values M, M1, M2, and aperiodic local maximum value M3 is committed to finding. Meanwhile, when another vehicle, etc. appears on the front side, the periodic stationary objects are locked by another vehicle, etc., and the counter values of the periodic local maximum values M, M1, M2 tend to be small.

[0063] Therefore, when the average value of the local maximum values M, M1, M2, having according to the definition of the periodicity in the current processing, not less than the average of the local maximum values, in accordance with the definition of frequency in the previous processing, to a predetermined value or more, another vehicle, etc. is on the opposite side of the periodic stationary objects, while considering the vehicle V cannot change lanes. Therefore, it is not necessary to detect a moving object, for example another vehicle. Meanwhile, when the average value of the local maximum values M, M1, M2, having according to the definition of the periodicity in the current processing, with less�one value of the local maximum values, having according to the definition of the periodicity in the previous processing, to a predetermined value or more, another vehicle, etc. is present before the periodic stationary objects, while considering the vehicle V can change the lane. Consequently, the detected moving object.

[0064] Thus, the detection system 1 of the periodic stationary objects and a method for the detection of periodic stationary objects in accordance with this embodiment of the allow a proper determination in accordance with the actual phenomenon.

[0065] it Should be noted that in the above embodiment, the image captured at the current time and during one time segment before, converted into a kind from height of bird's flight, and difference image PDtis created by performing the alignment for the converted species from the height of bird flight. Nevertheless, the present invention is not limited to this. For example, only the image during one time segment before converted to the view from the bird's eye view; a transformed view from height of bird's flight is subjected to registration, and then converted back into the original captured image; this image and the image in the current�the time can be used to create a differential image. In other words, the views from the bird's eye doesn't always have to be created explicitly provided that the image in the current time and the image during one time segment earlier are combined to create a difference image PDtout of the difference between such combined two images.

[0066] the Second variant of implementation

Further in this document based on the drawings described second variant implementation of the present invention. It should be noted that what is equivalent to that described in the first variant of implementation, are denoted by identical reference numbers, and description of the said falls.

[0067] Fig.11 is a schematic diagram of the system configuration 2 detection of periodic stationary objects according to the present variant implementation. This version of the implementation describes an example in which the detection system 2 of the periodic stationary objects mounted on the vehicle V. As shown in Fig.11, system 2 detection of periodic stationary objects includes a camera module 10 and 40 of the calculation.

[0068] Fig.12 is a view for illustrating the capture range of the camera 10 shown in Fig.11, etc. As shown in Fig.12, the camera 10 is made possible with�TEW to capture the image region at the rear side of the vehicle V under a predefined angle a review similarly to the first variant implementation. The angle a of the camera 10 is set so that the capture range of the camera 10 may include an adjacent lane or the side of the road in addition to the lane in which you are considering moving vehicle V.

[0069] the Module 40 performs various calculation processing for fragments in the regions A1, A2the detection of periodic stationary objects in the captured image captured by the camera 10. Because of this, the module 40 calculation determines whether or not three-dimensional objects present in the regions A1, A2definitions of periodic stationary objects. Area A1, A2definitions have a rectangular shape when viewed from above. The positions of the regions A1, A2definitions can be specified based on the relative positions of the vehicle V or can be set based on the position of a white strip of road markings through the use of the existing face recognition technology white strips of road marking, etc. the shape of the regions A1, A2definitions for image types with the bird's eye view is not limited to a rectangular shape. When the domain is a rectangular region in the actual space, form fields Asub> 1, A2definitions for image types from the height of bird flight can be trapezoidal form.

[0070] the Side regions A1, A2definition close to the vehicle V (the side along the direction of motion) are defined as reference line L1, L2contact with the ground. Reference line L1, L2ground contact means lines, which are in contact with the ground another vehicle VOpresent on the lane adjacent to the lane in which you are considering moving the vehicle V, and the periodic stationary object present along the roadside.

[0071] the Distance in the direction of movement of the vehicle from the rear boundary of a fragment of the vehicle V to the front boundary fragments of the regions A1, A2the definition is defined in such a way that at least the area A1, A2the definitions are within the angle a of the camera 10.

In addition, the length of each of the regions A1, A2determine in the direction of movement of the vehicle and its width in the direction orthogonal to the direction of movement of the vehicle is determined on the basis of the amount of the periodic stationary object that is to be detected. In this�present variant implementation, to distinguish the periodic stationary object from other vehicle VOthe length in the direction of movement of the vehicle is specified as the length, which can include, at least, another vehicle VO. In addition, the width in the direction orthogonal to the direction of movement of the vehicle, has a length, not including lanes (i.e., the second adjacent lanes), which are further adjacent to the traffic lanes adjacent to the right and left side in the picture of the view from the height of bird's flight.

[0072] Fig.13 is a flowchart for illustrating details of the module 40 of the calculation shown in Fig.11. As shown in Fig.13, the module 40 calculation includes a module 41 conversion point of view, the module 42 calculating the distribution of the edges, and the audit module 43, the module 44 of detecting variants of the periodic stationary objects, and the module 45 assessment of periodic stationary objects. It should be noted that the module 40 computing is a computer composed of CPU, RAM, ROM, etc. the Module 40 performs calculation processing of images, etc. according to the program set in advance to thereby implement the function of each module, such as module 41 conversion point of view, the module 42 calculating the distribution of the edges, counting m�Dul 43, module 44 detect variants of periodic stationary objects, and the module 45 assessment of periodic stationary objects.

[0073] the Module 41 the transformation of a viewpoint arranged to receive the captured image data in a predetermined region obtained through the capture by the camera 10. Module 41 the transformation of a viewpoint arranged to convert the viewpoint taken to convert captured image data into image data of a kind from height of the bird's flight in a state of viewing from a bird's eye view. Review status with the bird's eye view means a view from the virtual camera, which is directed, for example, vertically downward (or slightly angled down) in the direction from the sky.

[0074] Fig.14 is a view for illustrating detailed operations of the module 42 calculating the distribution of the edges, the calculating module 43, the module 44 of detecting variants of the periodic stationary objects and module 45 assessment of periodic stationary objects. It should be noted that although Fig.14 gives a description, by illustrating only the right side in the movement direction of the vehicle, which includes area a1definition module 42 calculating the distribution of the edges, and the audit module 43, modulating� 44 detect variants of periodic stationary objects, and the module 45 assessment of periodic stationary objects perform an identical processing for the area on the left side in the movement direction of the vehicle, includes area a2definitions.

[0075] As shown in Fig.13, the module 42 calculating the distribution of the edges includes a module 42a extraction of boundary elements and the module 42b calculate the waveform distribution of the edges. Module 42a extraction of boundary elements is arranged to calculate a chrominance difference image data in the form of aerial flight, obtained through conversion of the viewpoint via module 41 conversion point of view, to detect regional constituent element (hereinafter in this document called boundary element (characteristic point)) periodic stationary object included in the picture of the view from the height of bird flight. Module 42a extraction of boundary elements calculates the luminance difference between two pixels next to each of the multiple positions along a vertical virtual lines extending in the vertical direction in the actual space.

[0076] In particular, the module 42a extraction of boundary elements specifies the first vertical virtual line corresponding to the line segment that runs in a vertical direction in the actual space, and the second vertical virtual line corresponding to the line segment that runs in a vertical direction in the actual space for the image view from the heights of Petit�d flight after conversion point of view. Module 42a extraction of boundary elements sequentially receives the luminance differences along the first vertical virtual line between the points on the first vertical virtual lines and points on the second vertical virtual line.

[0077] the Module 42b calculate the waveform distribution of the edges is arranged to sum the number of boundary elements extracted by the module 42a extraction of boundary elements for each of the multiple vertical virtual lines and calculate the signal form of the distribution of edges based on the number of such aggregated boundary elements.

[0078] More detail about the operation of the module 42a extraction of boundary elements and module 42b calculate the waveform distribution of the edges.

As shown in Fig.14, the module 42a extraction of boundary elements of sets are some of the first vertical virtual lines Lai(hereinafter called the warning lines Lai), which are segments of lines extending in the vertical direction from points on the reference line L1ground contact in the actual space, and pass through the area A1definition. The number of warning lines Lainot limited to a particular manner. The following description describes the case when the set n warning lines Lai(i=1 to n).

[007] in addition, module 42a extraction of boundary elements of several sets of the second vertical virtual lines Lri(hereinafter called reference lines Lri), which properly match several warning lines Laiare segments of lines extending in the vertical direction from points on the reference line L1ground contact in the actual space, and pass through the area A1definitions..Each of the reference lines Lriis set to the position distant from the warning lines Laicorresponding to the actual space, at a predetermined distance (e.g. 10 cm). It should be noted that the lines corresponding to segments of lines extending in the vertical direction in the actual space, become lines radiating radially from the position of PSin the camera 10 in the picture of the view from the height of bird's flight.

[0080] Then the module 42a extraction of boundary elements of sets a few warning points Pajeach of the warning lines Lai. In the example shown in Fig.14, set warning points Pa1-Pa8but the number of warning points Pajnot limited to a particular manner. The following description describes the case when the set k warning points Paj (j=1-k).

[0081] Additionally, the module 42a extraction of boundary elements of several sets of reference points Prjon each of the reference lines Lriand the control points Prjproperly correspond to the red points Paj. Warning points Pajand the control points Prjcorresponding to each other are set at almost the same height in the actual space. It should be noted that the warning points Pajand the control points Prjnot necessarily always have to be at the identical height in the strict sense. Obviously, some height difference permissible provided that the height of the warning points Pajand control points Prjcan be regarded as identical.

[0082] the Module 42a extraction of boundary elements sequentially receives the luminance differences along each of the warning lines Laibetween warning points Pajand control points Prjcorresponding to each other. In the example shown in Fig.14, the module 42a extraction of boundary elements calculates the luminance difference between the first warning point Pa1and the first reference points Pr1and calculates the luminance difference between the second warning points Pa2and second control points Pr2. Then, similarly, the module 42a extraction of boundary elements�coefficients sequentially receives the luminance difference between the third to eighth warning points P a3-Pa8and third-eighth of the reference points Pr3-Pr8.

[0083] When the luminance difference between a warning point Pajand the reference point Prjis equal to or exceeds a predefined value, the module 42a extraction of boundary elements of what defines a boundary element is present between the warning point Pajand the reference point Prj. Thus, the module 42a extraction of boundary elements acts as a module of extracting feature points, configured to extract a boundary element (characteristic point) is present along each of multiple vertical virtual lines extending in the vertical direction in the actual space. In other words, the module 42a extraction of boundary elements extracts the characteristic point (boundary element) three-dimensional object from image data in a predetermined region (region definition) image view with the bird's eye view for each of the multiple sub-regions (each region near multiple vertical virtual lines) included in the predetermined region.

[0084] the Module 42b calculate the waveform distribution of the edges counts the number of boundary elements extracted by the module 42a extraction of boundary elements, there is a boat ride along� one warning lines L ai. Module 42b calculate the shape of the signal distribution preserves edges counted the number of such boundary elements as an attribute of each of the warning lines Lai.

[0085] the Module 42b calculate the shape of the signal distribution counts edges boundary elements for all warning lines Lai.It should be noted that the length of fragments warning lines Laioverlapping area A1the definitions differ from each other depending on where each of the warning lines Lai. The number counted of boundary elements can be normalized by dividing the number by the length overlapping fragment corresponding warning lines Lai.

[0086] In the example shown in Fig.14, another vehicle VOis displayed in A1definition. Assume that the warning line Laiset on a rubber slice the tires of the other vehicle VOand the reference line Lriis set equal to the position at a distance of approximately 10 cm from it on the wheel tire. In this case, since the first warning point Pa1and the first reference point Pr1are identical to the fragment of the tire, the luminance difference between them is small. Meanwhile, the second-eighth warning�incorporate the points P a2-Pa8are the rubber part of the tire, and the second to eighth control points Pr2-Pr8located in the fragment of a wheel tyre, which makes the luminance of the big differences between them. When the luminance difference between the second to eighth warning points Pa2-Pa8and the second one-eighth of the reference points Pr2-Pr8equal to or exceeds a predefined value, the module 42a extraction of boundary elements detects that boundary elements are present between the second to eighth warning points Pa2-Pa8and the second one-eighth of the reference points Pr2-Pr8. Additionally, since the second number is the eight warning points Pa2-Pa8present warning along the lines of Laiis 7, the module 42a extraction of boundary elements detects boundary elements 7 times. In this case, the module 42b calculate the waveform distribution of the edges calculates the marginal value of element as "7".

[0087] Additionally, the module 42b calculate the waveform of the distribution of edges forms a graph counter boundary element obtained for each warning lines Laiand receives the signal form of the distribution of the edges (data waveform). In particular, the module 42b calculate the waveform distribution of the edges illustrates the value of the account�ICA regional element in the plane thus the vertical axis represents the value of the counter edge of the element and the horizontal axis represents the position of the warning line Laion reference line L1ground contact in the actual space. If the warning line La1-Lanare set at regular intervals on the reference line L1ground contact in the actual space, the shape of the signal distribution of the edges can be obtained by simple combination of the counter values of boundary elements obtained for each warning lines Laiin order warning lines La1-Lan. In the example shown in Fig.14, warning line Laiset in the rubber part of the tires of the other vehicle VOhas the value of the counter edge item "7" in position intersecting the reference line L1contact with the ground in the picture of the view from the height of bird's flight.

[0088] Thus, the module 42b calculate the waveform distribution of the edges acts as a module to calculate the data waveform is arranged to sum the number of boundary elements extracted by the module 42a extraction of boundary elements for each of the multiple vertical virtual lines extending in the vertical direction in the actual space, � to calculate the frequency distribution of the edges (data waveform) based on the number of such aggregated boundary elements. In other words, the module 42b calculate the shape of the signal distribution computes the edges of the data waveform (based on the relative mutual arrangement of boundary elements) corresponding to the distribution of feature points (boundary elements) in the previously defined region (domain) in the picture of the view from the height of bird's flight.

[0089] the counting module 43 is arranged to detect the peaks of the waveform distribution of the edges computed by the module 42b calculate the waveform edges of the distribution module 42 calculating the distribution of the edges. The peak is the point in the waveform distribution of the edges in which the value of the counter edge of the element changes from increasing to decreasing. Counter module 43 detects the peaks after processing reduction of noise is performed for the waveform distribution of the edges, for example, by using a lowpass filter, a filter based on a moving average, etc. Here, the peak that needs to be detected, may be only a peak having a value equal to or greater than a predetermined threshold value. The pre-defined threshold value may be set equal to the value of 60% of the maximum value of the waveform distribution of the edges.

[0090] in addition, the audit module 43 for�counts the number of peaks, combined regular intervals, for peaks detected in this way (information peaks). In particular, the counting module 43 calculates the distance between the peaks found in this way, remove peaks, having calculated the distance between peaks in a predefined range, and counts the number of peaks. The "predetermined range" for the distance between the peaks may be a fixed value preset according to the type of periodic stationary objects that should be detected, or may be a variable value that is set based on the distances between the peaks consistently detected within a predetermined time period or more. It should be noted that when determining the peaks of the waveform distribution of the edges of the counting module 43 transmits the peak detected in some cases. In this case, the interval of peaks is detected as having a size, in two, three or more times larger than the actual interval. Therefore, to prevent incorrect counting the peaks, combined regular intervals, the "predetermined range" is set so that it includes the values corresponding to multiples of the spacing of the peaks, which should be extracted first. NRA�emer, when the interval of peaks that must be learned in the first place, is X, the "predetermined range" for the distance between the peaks is given as X±10%, 2X±20%, 3X±30%. Thus, the counting module 43 acts as a detection module information peaks, is arranged to detect the peaks of the data waveform.

[0091] the Module 44 of detecting variants of the periodic stationary objects is arranged to determine whether or not a three-dimensional object having the extracted boundary element option periodic stationary object, on the basis equals or exceeds or not the number of peaks (information peaks, calculated by the calculating module 43, a pre-defined threshold value Th2. In particular, when the number of peaks counted by the counting module 43, is equal to or exceeds a predetermined threshold value Th2module 44 detect variants of periodic stationary objects determines what objects corresponding to the counted peaks, are variants of periodic stationary objects. The threshold value Th2is a value determined according to the type of periodic stationary objects that must be detected, such as, for example, supports line electropure�Ah, road signs or telephone poles, and can be determined through experiments, etc. In particular, the threshold value Th2is set equal to, for example, the value of 3-100 (inclusive).

[0092] When a variant of the periodic stationary object is continuously detected within a predetermined time period, the module 45 assessment of periodic stationary objects is arranged to determine that a variant of the periodic stationary object is a periodic stationary object. In particular, when the module 44 of detecting variants of the periodic stationary objects continuously detects for a predetermined period of time, the condition in which the number of peaks is equal to or exceeds a predetermined threshold value Th2module 45 assessment of periodic stationary objects assesses that there is a high enough probability that the detected version of the periodic stationary object is a periodic stationary object. Then the module 45 assessment of periodic stationary objects determines what objects corresponding to the counted peaks are periodic stationary objects. "Predetermined time period" is the value defined soglasnoooo periodic stationary objects, to be detected, and may be obtained through experiments, etc. Value can be a fixed value or may vary according to the interval of capture of the camera 10 and the speed of movement of the vehicle V. In particular, the "predetermined time period" is set to, for example, 0.1-5 seconds.

[0093] the following describes a method for the detection of periodic stationary objects according to the present variant implementation. Fig.15 and 16 are block diagrams of the sequence of operations of way to illustrate details of the method of detection periodic stationary objects according to the present variant implementation. It should be noted that in Fig.15 and 16 describes the treatment aimed at the area A1definitions for convenience, however, identical treatment may also be performed for the area A2definitions.

[0094] As shown in Fig.15, first, in step S31, the module 41 conversion point of view takes the captured image data in a predetermined region obtained through the capture by the camera 10, and performs a transformation point of view for them to create image data of a kind from height of bird's flight.

[0095] Then, in step S32, the module 42 calculating the distribution of edges for�AET n warning lines L aion the area A1definitions and sets n reference lines Lrithat properly correspond to n warning lines Lai(i=1 to n). Module 42 calculating the distribution of edges specifies the segments of lines extending in the vertical direction from points on the reference line L1ground contact in the actual space, as warning lines Lai. In addition, the module 42 calculating the distribution of edges specifies the line segments that go in a vertical direction from points on the reference line L1ground contact in the actual space and spaced from the warning lines Laicorresponding to the actual space, on a pre-specified distance, as the reference lines Lri.

[0096] Then, in step S33, the module 42 calculating the distribution of the sets of edges k warning points Pajeach of the warning lines Laiand sets k control points Prj(j=1 to k) for each of the reference lines Lriand k control points Prjproperly correspond to the red points Paj. Module 42 calculating the distribution of the edges performs setting so that the warning points Pajand the control points Prjmatch virtually identical height in the actual space.

[0097] Then, �and step S34, module 42 calculating the distribution of edges is equal to or exceeds or not the luminance difference between a warning point Pajand the reference point Prjcorresponding to each other, predefined value. When determining that the luminance difference is equal to or exceeds a predefined value, the module 42 calculating the distribution of the edges determines that the boundary element is present between the warning point Pajand control points Prjdefined thus, and substitutes "1" for the value of the counter (bincount(i)) i-the warning lines Laiin step S35. In step S34, when determining that the luminance difference is less than a predetermined value, the module 42 calculating the distribution of the edges determines that the boundary element is not present between the warning point Pajand the reference point Prjdefined thus, and transfers the processing to step S36.

[0098] In step S36, the module 42 calculating the distribution of the edges determines whether the processing in step S34 warning for all points Pajon warning lines Laiprocessed at the moment. When determining that the processing in step S34 is not made for all warning points Pajthe module 42 to calculate the distribution of�AEB returns the processing to step S34, obtains a luminance difference between following warning point Paj+1and the reference point Prj+1what defines equals or exceeds or not the luminance difference of a predefined value. Thus, the module 42 calculating the distribution of the edges sequentially receives the luminance differences between warning points Pajand control points Prjalong the warning line Laiin the sequence. When the obtained luminance difference becomes equal to or greater than a predefined value, the module 42 calculating the distribution of the edges determines that there is a regional element.

[0099] After substituting "1" for the value of the counter (bincount(i)) i-the warning lines Laiin step S35, the module 42 calculating the distribution of the edges passes the processing to step S37. Then the module 42 calculating the distribution of the edges receives the luminance difference between the following warning point Paj+1and the reference point Prj+1what defines equals or exceeds or not the luminance difference of a predefined value. When determining that the luminance difference is equal to or exceeds a predefined value, the module 42 calculating the distribution of the edges determines that the boundary element is present between the warning point Paj1 and the reference point Prj+1defined thus, and increments a counter (bincount(i)) i-the warning lines Laiin step S38.

[0100] In step S37, when determining that the luminance difference is less than a predetermined value, the module 42 calculating the distribution of the edges determines that the boundary element is not present between the warning point Paj+1and control points Prj+1defined thus, and transfers the processing to step S39, skipping the step S38.

[0101] Then, in step S39, the module 42 calculating the distribution of the edges determines whether the processing in step S34 or S37 warning for all points Pajon warning lines Laiprocessed at the moment. When determining that the processing is not performed for all warning points Pajmodule 42 calculating the distribution of edges returns the processing to step S37, obtains a luminance difference between following warning point Paj+1and the reference point Prj+1what defines equals or exceeds or not the luminance difference of a predefined value. In step S39, when determining that the processing is executed for all warning points Pajmodule 42 calculating the distribution of the edges passes the processing to step S41. Thus, the module 42 computations�t distribution counts edges, how many boundary elements is present along the identical warning lines Laiand stores counted the number of such boundary elements as an attribute (bincount(i)) warning lines Lai.

[0102] it Should be noted that in step S36, when determining that the processing in step S34 is executed to all warning points Pajmodule 42 calculating the distribution of the edges determines that the boundary element not present on the warning lines Laiprocessed at the moment. Then the module 42 calculating the distribution of the edges substitutes "0" for bincount(i) in step S40, and passes the processing to step S41.

[0103] Then, in step S41, the module 42 calculating the distribution of the edges determines whether the above processing for all n warning lines Lai. When determining that the processing is not performed for all warning lines Laimodule 42 calculating the distribution of edges returns the processing to step S34, and performs processing for the following warning lines Lai+1. In step S41, when determining that the processing is executed for all warning lines Laimodule 42 calculating the distribution of the edges passes the processing to step S42.

[0104] Then, in step S42, the module 42 calculating the distribution of the edges forms a graph counter value bincount(i) (i=1 to n) cu�left element, obtained for each of n warning lines Laiand receives the signal form of the distribution of the edges. In particular, the module 42 calculating the distribution of the edges illustrates the value of the counter bincount(i) (i=1 to n) boundary of the element on the plane, while the vertical axis represents the count value of the regional element, and the horizontal axis represents the position of the warning line Laion reference line L1ground contact in the actual space.

[0105] Then, in step S43, the counting module 43 detects the peaks of the waveform distribution of the edges computed by the module 42 calculating the distribution of the edges.

[0106] Then, in step S44, the counting module 43 calculates the distance between the peaks found in this way.

[0107] Then, in step S45, the counting module 43 extracts the peaks, having calculated the distance between peaks in a predefined range, and counts the number of peaks.

[0108] Then, in step S46, the module 44 of detecting variants of the periodic stationary objects is equal to or exceeds or not the number of peaks counted by the counting module 43, a pre-defined threshold value Th2. When determining that the number of peaks is equal to or exceeds a predetermined threshold value Th2module 44 detection variantbinding stationary objects determines the the objects corresponding to the counted peaks, are variants of periodic stationary objects, and passes the processing to step S47.

[0109] In step S47, the module 45 assessment of periodic stationary objects evaluates the detected or it is not consistently the state in which the number of peaks is equal to or exceeds a predetermined threshold value Th2pre-specified number of times or more. When assessing what the state in which the number of peaks is equal to or exceeds a predetermined threshold value Th2successively detected a predetermined number of times or more, the module 45 assessment of periodic stationary objects assesses that the objects corresponding to the counted peaks are periodic stationary objects, and substitutes "1" for the flag f_shuki in step S48. Meanwhile, in step S47, when assessing what the state in which the number of peaks is equal to or exceeds a predetermined threshold value Th2not detected successively a predetermined number of times or more, the module 45 assessment of periodic stationary objects skips the step S48 and supports the value of the flag f_shuki. After this is completed, the processing in Fig.15 and 16.

[0110] In step S46, when determining that the number of peaks is less�e pre-defined threshold value Th 2module 44 detect variants of periodic stationary objects transfers the processing to step S49.

[0111] In step S49, the module 45 assessment of periodic stationary objects evaluates the detected or it is not consistently the state in which the number of peaks is less than a predetermined threshold value Th2pre-specified number of times or more. When assessing what the state in which the number of peaks is less than a predetermined threshold value Th2successively detected a predetermined number of times or more, the module 45 assessment of periodic stationary objects assesses that the objects corresponding to the counted peaks are periodic stationary objects, and substitutes "0" for the flag f_shuki in step S50. Meanwhile, in step S49, when assessing what the state in which the number of peaks is less than a predetermined threshold value Th2not detected successively a predetermined number of times or more, the module 45 assessment of periodic stationary objects skips the step S50 and supports the value of the flag f_shuki. After this is completed, the processing in Fig.15 and 16.

[0112] the system 2 detection of periodic stationary objects and the method of detection periodic stationary object� according to this embodiment of the boundary element (characteristic point) of the three-dimensional object is extracted from image data in a predetermined area of an image with height of bird's flight for each of the areas (sub-areas) adjacent with multiple vertical virtual lines included in the predetermined region; calculates the waveform distribution of the edges (data waveform) corresponding to the distribution of boundary elements in a predetermined area; and whether or not the three-dimensional object having the extracted boundary element option periodic stationary object is determined on the basis of equal to or exceeds or not the number of peaks (information peaks) of the waveform distribution of the edges of the pre-defined threshold value. Similarly to the first variant implementation, it provides a more precise extraction of the periodicity (frequency) of periodic stationary objects as information peaks form data signal, and variants of the periodic stationary objects can easier be extracted from three-dimensional objects included in the captured image. Therefore, accurate extraction of periodic stationary objects.

[0113] in addition, in the system 2 detection of periodic stationary objects and the detection of periodic stationary objects in accordance with this embodiment of the summed number of boundary elements present along each of several vertically�x virtual lines, running in the vertical direction in the actual space, and turns the signal form of the distribution of edges based on the number of such aggregated boundary elements. In addition, when the number of peaks of the waveform distribution of the edges is equal to or exceeds a predetermined threshold value Th2determined that a three-dimensional object having the extracted boundary element, is a variant of the periodic stationary object. Consequently, reliably detected case in which edge running in the vertical direction, is very tightly integrated without determining that certain three-dimensional objects are stationary objects or moving objects. Thus, there is a possibility it is easier to detect variants of periodic stationary objects, which are more likely are periodic stationary objects.

[0114] In particular, the system 2 detection of periodic stationary objects and the detection of periodic stationary objects according to this variant implementation, the number of peaks, aligned at regular intervals, is calculated for the peaks of the waveform distribution of the edges. Consequently, it can more reliably detect variants of periodic stationary objects that have edges going in the Vert�local direction and combined with high density at regular intervals, and more likely are periodic stationary objects.

[0115] Additionally, the system 2 detection of periodic stationary objects and the detection of periodic stationary objects according to this variant implementation, when the option of periodic stationary object is continuously detected within a predetermined period of time, determined by what version of the periodic stationary object is a periodic stationary object. Consequently, prevents erroneous determination due to noise, and can be more reliably detected periodic stationary objects.

[0116] the Third variant of implementation

Further in this document based on the drawings described third embodiment of the present invention. It should be noted that what is equivalent to that described in the first and second embodiments, are denoted by identical reference numbers, and description of the said falls.

[0117] the detection System 3 of the periodic stationary objects according to the present variant implementation has a schematic configuration identical to the schematic system configuration of the 1 detection of periodic stationary objects, shown in Fig.1, but includes a module 40' �of ycycline module instead of 30 calculations. In particular, the detection system 3 of the periodic stationary objects according to the present variant implementation includes a chamber 10, a sensor 20 of the vehicle speed and the module 40' calculations.

[0118] Fig.17 is a flowchart for illustrating details of the module 40' calculations according to this variant implementation. As shown in Fig.17, the module 40' calculations includes a module 41 conversion point of view, the module 42 calculating the distribution of the edges, and the audit module 43', module 44 detect variants of periodic stationary objects, the module 51 combining module 52 for calculating the difference between the module 53 of the assessment of the periodic stationary objects. It should be noted that the module 40' computing is a computer composed of CPU, RAM, ROM, etc. the Module 40' calculations performs image processing, etc. according to the program set in advance to thereby implement the function of each module, such as module 41 conversion point of view, the module 42 calculating the distribution of the edges, and the audit module 43', module 44 detect variants of periodic stationary objects, the module 51 combining module 52 for calculating the difference between the module 53 of the assessment of the periodic stationary objects.

[0119] the counting module 43' according to this variant implementation is made with the ability to detect peak� waveform distribution of the edges, calculated by the module 42 calculating the distribution of the edges, and count the number of peaks. Counter module 43' differs from the counting module 43 according to the second embodiment of the fact that the counting module 43' counts the number of peaks without exception peaks of the distance between the peaks outside a predetermined range.

[0120] the Module 44 of detecting variants of the periodic stationary objects is arranged to determine whether or not the objects corresponding peaks, versions of the periodic stationary objects, on the basis equals or exceeds or not the number of peaks (information peaks, calculated by the calculating module 43', a pre-defined threshold value Th3. In particular, when the number of peaks counted by the counting module 43', equal to or greater than a predetermined threshold value Th3module 44 detect variants of periodic stationary objects determines what objects corresponding peaks, are variants of periodic stationary objects. The threshold value Th3is a value determined according to the type of periodic stationary objects that must be detected, such as, for example, power pylons, road signs or Telegraph St�foreheads, and can be obtained through experiments, etc. In particular, the threshold value Th3is set equal to, for example, the value of 3-100 (inclusive).

[0121] Fig.18 shows graphs for illustrating detailed operations of the module 51 of alignment. Module 51 of the alignment is made with the ability to consistently take the shape of the signal distribution of the edges computed by the module 42 calculating the distribution of the edges, and align the position of the received waveform distribution of the edges at different points in time based on the speed of the vehicle V is determined by means of the sensor 20 of the vehicle speed. For example, assume that the module 51 of the alignment takes the form Et- ∆ Tsignal distribution of the edges computed by the module 42 calculating the distribution of edges at time t-Δt (second time), and Etsignal distribution edges, computed at time t (the first time). Furthermore, suppose that as soon as the vehicle V moves during one time segment (Δt), the signal form of the distribution moves to the brim δ relative to the coordinate system. In this case, the module 51 of combining shifts the Etsignal distribution edges by δ along the horizontal axis, as shown in Fig.18, to thus�m to negotiate a position of the form E tsignal distribution edges with the position of the form Et- ∆ Tsignal distribution edges. Thus, the module 51 receives a combination Et'signal distribution edges. Here, the combination of the positions of the waveforms of the distribution of the edges means that warning when line Laicorresponding to the point (for example, G1) in one form of the signal distribution of the edges, and warning line Laicorresponding to the point (e.g., G2) in another form of the signal distribution of the edges that are present in identical or approximately identical to the position in the actual space, form(s) signal distribution of the edges is moved in parallel so that the values of the horizontal coordinates of points G1 and G2 coincide with each other. In this regard, the duration of one time segment (Δt) may be, for example, a predetermined time period, such as one cycle management, and may be an arbitrary period of time.

[0122] Fig.19 shows graphs for illustrating detailed operations of the module 52 for calculating the difference between. The module 52 for calculating the difference between arranged to take the form of Et- ∆ Tsignal distribution of edges and Et'the signal distribution of the edges computed by the module 51 of alignment, and calculate the distribution of the absolute values of the difference �between them |E t- ∆ T-Et'|. When the waveform of the distribution of edges is extracted from the periodic stationary object, Et- ∆ Tsignal distribution edges optimally coincides with Et'signal distribution edges. Accordingly, the absolute value of the difference |Et- ∆ T-Et'|, in General, are small values, as shown in Fig.19(a). Meanwhile, when the waveform of the distribution of edges is extracted from moving objects, Et- ∆ Tsignal distribution edgesnot the same as the Et'the signal distribution of the edges, and the absolute value of the difference |Et- ∆ T-Et'| significantly changed compared with Fig.19(a), as shown in Fig.19(b).

[0123] the Module 53 of the assessment of the periodic stationary objects is arranged to integrate the absolute value |Et- ∆ T-Et'|calculated by the module 52 for calculating the difference between to calculate the integrated value of ID1(the first integrated value) and to calculate the integrated value of I1(second integrated value) of the form Et'signal distribution edges. Additionally, the module 53 of the assessment of the periodic stationary objects is arranged to calculate the ratio of the integrated value of ID1toward an integrated value of I1(ID1/I1and op�edalati, is or is not a variant of the periodic stationary object detected by the module 44 of detecting variants of the periodic stationary objects, stationary (stationarity), on the basis of less or no value of the ratio of a predetermined threshold value Th4. The threshold value Th4is a value determined according to the type of periodic stationary objects that must be detected, such as, for example, power pylons, road signs or telephone poles, and can be obtained through experiments, etc. When the ratio of the integrated value of ID1toward an integrated value of I1(ID1/I1) is less than a predetermined threshold value Th4the module 53 of the assessment of the periodic stationary objects determines what version of the periodic stationary object is stationary.

[0124] When the stationary version of the periodic stationary object is continuously detected within a predetermined time period, the module 53 of the assessment of the periodic stationary objects determines what version of the periodic stationary object is a periodic stationary object. In particular, when the state in which the ratio of ID1/I1less than a predetermined threshold value Th 4continuously detected within a predetermined time period, the module 53 of the assessment of the periodic stationary objects determines that there is a high enough probability that the detected version of the periodic stationary object is a periodic stationary object. Then the module 53 of the assessment of the periodic stationary objects determines what objects corresponding to the counted peaks are periodic stationary objects. "Predetermined time period" is a value defined according to the type of periodic stationary objects that should be detected, and may be obtained through experiments, etc. Value can be a fixed value or may vary according to the interval of capture of the camera 10 and the speed of movement of the vehicle V. In particular, the "predetermined time period" is set to, for example, 0.1-5 seconds. This time period can ensure the reliability of determining what versions of the periodic stationary objects are periodic stationary objects, and is less than the average interval in which an error occurs in the measurement of the speed of the vehicle V, etc.

[0125] Then describes�. a method of detecting periodic stationary objects according to the present variant implementation. Fig.20 is a block diagram of the sequence of operations of way to illustrate details of the method of detection periodic stationary objects in accordance with this embodiment of the corresponding to Fig.16 of the second variant implementation. It should be noted that since the processing of steps S31-S41 detection of periodic stationary objects according to the present variant implementation is identical to the processing of steps S31-S41 of the second embodiment, their illustration and description is omitted. In addition, further in this document describes the processing aimed at the area A1definitions for convenience; however, identical treatment may also be performed for the area A2definitions.

[0126] As shown in Fig.20, in step S51, running after step S41 (see Fig.15), the module 42 calculating the distribution of the edges forms a graph counter value bincount(i) (i=1 to n) boundary of the element, obtained for each of n warning lines Laiand receives the signal form of the distribution of the edges. In particular, the module 42 calculating the distribution of the edges illustrates the value of the counter bincount(i) (i=1 to n) boundary of the element on the plane, while the vertical axis represents the count value of the regional element, and the horizontal axis represents the position of warning�line L aion reference line L1ground contact in the actual space.

[0127] Then, in step S52, the counting module 43' detects the peaks of the waveform distribution of the edges computed by the module 42 calculating the distribution of the edges, and counts the number of peaks.

[0128] Then, in step S53, the module 44 of detecting variants of the periodic stationary objects is equal to or exceeds or not the number of peaks counted by the counting module 43', a pre-defined threshold value Th3. When determining that the number of peaks is equal to or exceeds a predetermined threshold value Th3module 44 detect variants of periodic stationary objects determines what objects corresponding peaks, are variants of periodic stationary objects, and passes the processing to step S54. In step S53, when determining that the number of peaks counted by the counting module 43', less than a predetermined threshold value Th3terminates the processing of Fig.20.

[0129] Then, in step S54, the module 51 of combining combines the positions of the waveforms of the distribution of edges within various time points taken from the module 42 calculating the distribution of the edges, based on the speed of the vehicle V, the Oprah�distributed by means of the sensor 20 of the vehicle speed. In particular, when the waveform of the distribution moves to the brim δ relative to the coordinate system of the graph as discussed, the vehicle V moves during one time segment (Δt), the module 51 of combining shifts the Etsignal distribution edges by δ along the horizontal axis and receives the Et'signal distribution edges.

[0130] Then, in step S55, the module 52 for calculating the difference calculates the distribution of the absolute values of the difference |Et- ∆ T-Et'| between Et- ∆ Tsignal distribution of edges and Et'the signal distribution of the edges computed by the module 51 of alignment.

[0131] Then, the module 53 of the assessment of the periodic stationary objects calculates an integrated value of I1form Et'the signal distribution of the edges in step S56, and calculates the integrated value of ID1absolute values |Et- ∆ T-Et'| on the next step S57.

[0132] Then, in step S58, the module 53 of the assessment of the periodic stationary objects, calculates the ratio (ID1/I1integrated values of ID1toward an integrated value of I1and determines whether or not a variant of the periodic stationary object detected by the module 44 detect variants periodic stationary about�projects, stationary, on the basis of less or no value of the ratio of a predetermined threshold value Th4. When determining that the ratio of ID1/I1less than a predetermined threshold value Th4the module 53 of the assessment of the periodic stationary objects determines what version of the periodic stationary object is stationary, and transfers the processing to step S59.

[0133] In step S59, the module 53 of the assessment of the periodic stationary objects is detected or not consistently state in which the ratio of ID1/I1less than a predetermined threshold value Th4pre-specified number of times or more, in other words, it is determined whether or not the state continuously for a predetermined period of time. When determining that the condition in which the ratio of ID1/I1less than a predetermined threshold value Th4successively detected a predetermined number of times or more, the module 53 of the assessment of the periodic stationary objects determines what objects corresponding to the counted peaks are periodic stationary objects, and substitutes "1" for the flag f_shuki in step S60. Meanwhile, in step S59, when determining t�, that the condition in which the ratio of ID1/I1less than a predetermined threshold value Th4not detected successively a predetermined number of times or more, the module 53 of the assessment of the periodic stationary objects skips S60 and supports the value of the flag f_shuki. Then terminates the processing in Fig.20.

[0134] In step S58, when determining that the ratio of ID1/I1is equal to or exceeds a predetermined threshold value Th4the module 53 of the assessment of the periodic stationary objects transfers the processing to step S61.

[0135] In step S61, the module 53 of the assessment of the periodic stationary objects is detected or not consistently state in which the ratio of ID1/I1is equal to or exceeds a predetermined threshold value Th4pre-specified number of times or more. When determining that the condition in which the ratio of ID1/I1is equal to or exceeds a predetermined threshold value Th4successively detected a predetermined number of times or more, the module 53 of the assessment of the periodic stationary objects determines what objects corresponding to the counted peaks are periodic stationary objects, and �otstavlyaet "0" for the flag f_shuki in step S62. Meanwhile, in step S61, when determining that the condition in which the ratio of ID1/I1is equal to or exceeds a predetermined threshold value Th4not detected successively a predetermined number of times or more, the module 53 of the assessment of the periodic stationary objects skips step S62 and supports the value of the flag f_shuki. Then terminates the processing in Fig.20.

[0136] In the system 3 detection of periodic stationary objects and the detection of periodic stationary objects in accordance with this embodiment of the boundary element (characteristic point) of the three-dimensional object is extracted from image data in a predetermined area of the image with the bird's eye view for each of the areas (sub-areas) adjacent with multiple vertical virtual lines included in the predetermined region; calculates the waveform distribution of the edges (data waveform) corresponding to the distribution of boundary elements in a predetermined area; and whether or not the three-dimensional object having the extracted boundary element, option periodic stationary object is determined on the basis of equal to or exceeds or not the number of peaks (information peaks) of the waveform distribution� edges of the pre-defined threshold value. Similar to the first embodiment of the second variant of implementation, it provides a more precise extraction of the periodicity (frequency) of periodic stationary objects as information peaks form data signal, and variants of the periodic stationary objects can easier be extracted from three-dimensional objects included in the captured image. Therefore, accurate extraction of periodic stationary objects.

[0137] in addition, in the system 3 detection of periodic stationary objects and the detection of periodic stationary objects according to this variant implementation, similar to the second variant of implementation, we summarize the number of boundary elements present along each of multiple vertical virtual lines extending in the vertical direction in the actual space, and turns the signal form of the distribution of edges based on the number of such aggregated boundary elements. In addition, when the number of peaks of the waveform distribution of the edges is equal to or exceeds a predetermined threshold value Th3determined that a three-dimensional object having the extracted boundary element, is a variant of the periodic stationary object. Therefore, similarly to the second embodiment of the, �Adina case detected, in which region, running in a vertical direction, is very tightly integrated without determining that certain three-dimensional objects are stationary objects or moving objects. Thus, there is a possibility it is easier to detect variants of periodic stationary objects, which are more likely are periodic stationary objects.

[0138] Additionally, in the system 3 detection of periodic stationary objects and the detection of periodic stationary objects according to the present embodiment of the position of the form Etsignal distribution edgesat time t coincides with the position of the form Et- ∆ Tsignal distribution of edges at time t-Δt on the basis of the speed of a moving object, and calculates a differential signal form of the distribution of |Et- ∆ T-Et'| between Et- ∆ Tsignal distribution edgesat time t-Δt and this combined form Et'signal distribution edges. Then integrate this differential waveform distribution |Et- ∆ T-Et'|to calculate the integrated value of ID1, combined and integrated Et'signal distribution edges to calculate the integrated value of I1. Calculates the ratio of the integrated value of ID1� integrated value of I 1(ID1/I1), and then whether or not it is a variant of the periodic stationary object stationary, is determined on the basis of less or no value of the ratio of a predetermined threshold value Th4. Therefore, it is possible to detect stationary versions of the periodic stationary objects, which are more likely are periodic stationary objects, and periodic stationary objects can be detected even easier and more reliable.

[0139] in addition, in the system 3 detection of periodic stationary objects and the detection of periodic stationary objects according to this variant implementation, when during a predetermined time period continuously detected what version of the periodic stationary object is stationary, it is determined that the option of a periodic stationary object is a periodic stationary object. Consequently, prevents erroneous determination due to noise, and can be more reliably detected periodic stationary objects.

[0140] a Modified example

In the above third embodiment, the implementation calculates the ratio of the integrated value of ID1toward an integrated value of I1(ID1/I1 ), and then whether or not it is a variant of the periodic stationary object stationary, is determined on the basis of less or no value of the ratio of a predetermined threshold value Th4. However, the method of determination is not limited to this.

[0141] In this modified example, the module 52 for calculating the difference between arranged to calculate the distribution of the absolute values of the difference (first difference) |(Et- ∆ T-Et'| between Et'signal distribution edges and Et- ∆ Tsignal distribution edgesand to calculate the distribution of the absolute values of the difference (second difference) |(Et- ∆ T-Et| between Etsignal distribution edgesand Et- ∆ Tsignal distribution edges.

[0142] in addition, the module 53 of the assessment of the periodic stationary objects is arranged to integrate the absolute value |Et- ∆ T-Et'|calculated by the module 52 for calculating the difference between to calculate the integrated value of ID1(the first integrated value), and to integrate the absolute value |Et- ∆ T-Et| to calculate the integrated value of ID2(second integrated value).

[0143] Additionally, the module 53 of the assessment of the periodic stationary objects are arranged to bicicletteria integrated value of I D1toward an integrated value of ID2(ID1/ID2) and to determine whether or not a variant of the periodic stationary object detected by the module 44 of detecting variants of the periodic stationary objects, stationary, on the basis of less or no value of the ratio of a predetermined threshold value Th5. The threshold value Th5is a value determined according to the type of periodic stationary objects that must be detected, such as, for example, power pylons, road signs or telephone poles, and can be obtained through experiments, etc. When the ratio of the integrated value of ID1toward an integrated value of ID2(ID1/ID2) is less than a predetermined threshold value Th5the module 53 of the assessment of the periodic stationary objects determines what version of the periodic stationary object is stationary.

[0144] When the stationary version of the periodic stationary object is continuously detected within a predetermined time period, the module 53 of the assessment of the periodic stationary objects determines what version of the periodic stationary object is a periodic stationary object. In h�particular, when the state in which the ratio of ID1/ID2less than a predetermined threshold value Th4continuously detected within a predetermined time period, the module 53 of the assessment of the periodic stationary objects determines that there is a high enough probability that the detected version of the periodic stationary object is a periodic stationary object. Then the module 53 of the assessment of the periodic stationary objects determines what objects corresponding to the counted peaks are periodic stationary objects.

[0145] the following describes a method for the detection of periodic stationary objects according to this modified example. Fig.21 is a block diagram of the sequence of operations of way to illustrate details of the method of detection periodic stationary objects according to this modified example corresponding to Fig.16 and 20. It should be noted that since the processing of steps S31-S41 detection of periodic stationary objects according to this modified example is identical to the processing of steps S31-S41 the above-described embodiment, their illustration and description is omitted. In addition, in the method of detection periodic stationary lens�tov according to this modified example, the processing, equivalent to the processing described in the second and third embodiments are indicated by identical reference numbers, and description is omitted.

[0146] In this modified example, as shown in Fig.21, in step S55', running after step S55, the module 53 of the assessment of the periodic stationary objects computes the distribution of the absolute values of the difference |Et- ∆ T-Et| between Et- ∆ Tsignal distribution of edges and Etsignal distribution of the edges computed by the module 51 of alignment.

[0147] Then, the module 53 of the assessment of the periodic stationary objects calculates an integrated value of ID1absolute values |Et- ∆ T-Et'| in step S57, and calculates the integrated value of ID2absolute values |Et- ∆ T-Et| in a subsequent step S57'.

[0148] Then, in step S58', the module 53 of the assessment of the periodic stationary objects computes the ratio of the integrated value of ID1toward an integrated value of ID2(ID1/ID2) and determines whether or not a variant of the periodic stationary object detected by the module 44 of detecting variants of the periodic stationary objects, stationary, on the basis of less or no value of the ratio of a predetermined threshold value h 5. When determining that the ratio of ID1/ID2less than a predetermined threshold value Th5the module 53 of the assessment of the periodic stationary objects determines what version of the periodic stationary object is stationary, and transfers the processing to step S59. Meanwhile, in step S58, when determining that the ID1/ID2is equal to or exceeds a predetermined threshold value Th5the module 53 of the assessment of the periodic stationary objects transfers the processing to step S61.

[0149] the Processing after step S59, and the processing after step S61 are identical to the processings in the third variant of implementation, and their description is omitted.

[0150] In this modified example calculates the first differential signal form of the distribution of |Et- ∆ T-Et'| between Et- ∆ Tsignal distribution of edges at time t-Δt and this combined form Et'signal distribution edges and a second differential signal form of the distribution of |Et- ∆ T-Et| between Etsignal distribution of edges at time t and Et- ∆ Tsignal distribution of edges at time t-Δt. Then integrates the first differential signal form of the distribution of |Et- ∆ T-Et'|to calculate the integrated value of ID1and integrates the second difference forms� signal distribution |E t- ∆ T-Et| to calculate the integrated value of ID2. Calculates the ratio of the integrated value of ID1toward an integrated value of ID2(ID1/ID2and that is an option or not periodic stationary object stationary, is determined on the basis of less or no value of the ratio of a predetermined threshold value Th5. Since the denominator of the ratio is the integrated value of ID2difference between Etsignal distribution of edges at time t and Et- ∆ Tsignal distribution of edges at time t-Δt, it makes more significant the difference between the value of the ratio obtained when the form EtEt- ∆ Tsignal distribution edges are extracted from moving objects, and value the relationship, obtained by extraction of periodic stationary objects, and can be more reliably detected stationary versions of the periodic stationary objects.

[0151] it Should be noted that in the third variant of implementation and the modified example that is compared with predefined threshold values when determining whether or not a variant of the periodic stationary object stationary is not limited to the relation ID1/I1and ID1/ID2 . The ratio may be, for example, by the ratio of the integrated value of ID1toward an integrated value of I2form Etsignal distribution edges (ID1/I2) or by the ratio of the integrated value of ID1toward an integrated value of I3form Et- ∆ Tsignal distribution edges, computed at time t-Δt (ID1/I3).

[0152] in addition, all of the integrated value of I1, I2, I3, ID1, ID2constituting the denominators and numerators of the above relations are obtained by integrating the waveform of the distribution of the edges or of the absolute values of the difference, but may be values obtained by integrating the waveform, which follows from squaring the values of the waveforms of the distribution of edges or their difference.

[0153] the Above-described embodiments of the present invention. However, these implementation options are just examples described in order to simplify the understanding of the present invention, and the present invention is not limited to variants of implementation. Scope of the present invention includes not only the specific technical matters disclosed in the above-described embodiments, but various modifications, variations, alternative technologies,etc., which can be easily developed.

[0154] for Example, in the above embodiments, the vehicle speed of the vehicle V is determined on the basis of the signal from the sensor 20 of the vehicle speed. However, without limiting this, the speed can be estimated from multiple images at different points of time. In this case, you no longer need the speed sensor of the vehicle, and can be simplified in configuration.

[0155] This application claims priority of patent application (Japan) No. 2011-034097, filed on February 21, 2011, which contains all of the content in this document by reference.

Industrial applicability

[0156] In the detection system of periodic stationary objects and the detection of periodic stationary objects according to the present invention, the characteristic point three-dimensional object is extracted from image data in a predetermined area of the image with the bird's eye view, respectively, for a number of subfields included in the predetermined region; calculated form data signal corresponding to the distribution of characteristic points in a predetermined region for displaying the view from the height of bird's flight; and that, whether or not the three-dimensional object having the extracted characteristic point, VA�Ianto periodic stationary object, is determined on the basis that equals or exceeds either no information peaks form data signal a predetermined threshold value. This provides a more precise extraction of the periodicity (frequency) of periodic stationary objects as information peaks form data signal, and variants of the periodic stationary objects can easier be extracted from three-dimensional objects included in the captured image. Therefore, accurate extraction of periodic stationary objects.

Number list of links

[0157] 1, 2, 3 - detection system of periodic stationary objects

10 - camera (capture device images)

20 - sensor vehicle speed (speed detector)

30, 40 - module calculations

31, 41 - conversion module point of view

32, 51 - combining module

33 - detection module of three-dimensional objects

34 is a module for computing variants of the magnitude of displacement

35, 43 and audit module

36 module calculating a range of motion

37, 45, 53 - module assessment periodic stationary objects

37a - detection module variants periodic stationary objects

37b - determining module periodicity

38 - detection module of the lane change (detection module transverse motion)

42 module you�of Elenia distribution edges

44 - detection module variants periodic stationary objects

52 module for calculating the difference between

a - angle of view

PBt- picture of the view from the height of bird flight

PDt- difference image

V - consider the vehicle

1. Detection system of periodic stationary objects to detect periodic stationary object in the vicinity of a moving object, containing:
device capturing images, mounted on a moving object and allowing the image capturing the surroundings of a moving object;
- conversion module viewpoint is arranged to convert at the point of view for the images that are captured by a capture device images to create a picture of the view from the height of bird's flight;
- the module of extracting feature points, configured to extract a characteristic point three-dimensional object from image data in a predetermined area of the image with the bird's eye view for each of the plurality of sub-regions included in the predetermined region;
- the module is calculating form data signal, configured to calculate data of the waveform corresponding to the distribution of feature points extracted RVBR�the rotary module of extracting feature points in a predetermined region of the image with the bird's eye view;
- detection module information peaks, is arranged to detect the peaks of the data signal form;
- detection module variants periodic stationary objects, performed with the opportunity to assess whether or not a three-dimensional object having the feature point extracted by the extraction module feature points, a variant of the periodic stationary object, on the basis that equals or exceeds either no information peaks predetermined first threshold value, and
module assessment periodic stationary objects, is arranged to determine that a variant of the periodic stationary object is a periodic stationary object when the periodic variant of the stationary object detected by the detection module variants periodic stationary objects, and the detection is carried out when the predetermined condition.

2. Detection system of periodic stationary objects according to claim 1, in which:
the extraction module extracts the characteristic points of the boundary elements along each of the multiple vertical virtual lines extending in the vertical direction in the actual space,
module calculate the data waveform summarizes the number of boundary e�cops extracted by the extraction module feature points for each of the multiple vertical virtual lines, and calculates the signal form of the distribution of edges based on the number of such aggregated boundary elements,
module detection information of the peaks is a counting module, configured to detect peaks of the waveform distribution of the edges calculated by the calculation module of the data waveform, and calculate the number of peaks found in this way, and
- detection module variants periodic stationary objects determines what three-dimensional object having the feature point extracted by the extraction module feature points, is a variant of the periodic stationary object when the number of peaks counted by the counting module is equal to or exceeds a predefined second threshold value.

3. Detection system of periodic stationary objects according to claim 2, in which the counting module detects the peaks of the waveform distribution of the edges calculated by the calculation module of the data waveform, and calculates the number of peaks, aligned at regular intervals, for peaks detected in this way.

4. Detection system of periodic stationary objects according to any of claims.1-3, in which the module is about�sister periodic stationary objects determines the what version of the periodic stationary object is a periodic stationary object when the option of periodic stationary object is continuously detected within a predetermined time period.

5. Detection system of periodic stationary objects according to claim 2, further comprising:
a velocity detector adapted to determine the speed of a moving object;
a combining module, configured to combine the position of the waveform distribution of the edges in the first time, calculated by the calculation module of the data waveform, the position of the waveform distribution of the edges in the second time, different from the first time, based on speed of movement defined by velocity detector, and
- module for calculating the difference between, configured to calculate a differential form of the signal distribution between the shape of the signal distribution of the edges in the second time and the shape of the signal distribution of the edges, combined by the combining module, wherein:
module assessment periodic stationary objects:
- integrates a differential waveform distribution calculated by the calculation module of the difference to calculate a first integrated value,
- integrates to any of the waveform �of raspredelenie edges at first, the shape of the signal distribution of the edges in the second time and waveform of the distribution of the edges, combined by the combining module to calculate a second integrated value,
- calculates the ratio of the first integrated value to the second integrated value and
- determines whether or not a variant of the periodic stationary object stationary, on the basis of less or no value of the ratio of a predetermined third threshold value.

6. Detection system of periodic stationary objects according to claim 2, further comprising:
a velocity detector adapted to determine the speed of a moving object;
a combining module, configured to combine the position of the waveform distribution of the edges in the first time, calculated by the calculation module of the data waveform, the position of the waveform distribution of the edges in the second time, different from the first time, based on speed of movement defined by velocity detector, and
- module for calculating the difference between, is arranged to calculate the first difference form of the signal distribution between the shape of the signal distribution of the edges in the second time and the shape of the signal distribution of the edges, combined by the combining module, and the WTO�th differential form of the signal distribution between the shape of the signal distribution of the edges in the first time and the shape of the signal distribution of the edges in the second time in this case:
module assessment periodic stationary objects:
- integrates the first differential waveform distribution to calculate a first integrated value,
- integrates the second difference signal waveform distribution to calculate a second integrated value,
- calculates the ratio of the first integrated value to the second integrated value and
- determines whether or not a variant of the periodic stationary object stationary, on the basis of less or no value of the ratio of a predetermined fourth threshold value.

7. Detection system of periodic stationary objects according to any of claims.5 and 6, in which the assessment module of the periodic stationary objects determines what version of the periodic stationary object is a periodic stationary object, with the continuous discovery that the option for a periodic stationary object is stationary for a predetermined period of time.

8. Detection system of periodic stationary objects according to claim 1, further comprising:
a velocity detector adapted to determine the speed of a moving object, and
- the module is calculating the range of movements made with the possibility of computing�you the range of motion of the periodic stationary objects for the image view from the height of bird flight based on the interval of capture devices capture images and speed of movement, determined by the velocity detector, in this case:
module extraction feature points includes:
a combining module, configured to combine the position of the image data at different points in time created by the module conversion points of the review;
- detection module of three-dimensional objects, is arranged to detect a variety of three-dimensional objects on the basis of data on the differential image of the image data at different points in time, combined by the combining module, and
- the module for computing variants of the magnitude of movements made with the possibility of options to calculate the magnitude of displacement of a plurality of three-dimensional objects detected by the detection module of three-dimensional objects,
module calculate the data waveform includes a counting module, configured to calculate the options value of the displacement calculated by the calculation module variants of displacement magnitude, and form a histogram as the data waveform,
- when the values of peaks of the histogram generated by the counting module is equal to or exceeds a predetermined fifth threshold value, the detection module variants periodic stationary objects determines what three-dimensional� objects associated with variations of the magnitude of the displacement, the respective peaks, are variants of periodic stationary objects and
module assessment periodic stationary objects assesses what options are periodic stationary objects are periodic stationary objects, when the peaks of the histogram are present in the range of displacement calculated by the calculation module of the range of motion.

9. Detection system of periodic stationary objects according to claim 8, in which the module for computing variants of displacement magnitude calculates many options of displacement magnitude for each of three-dimensional objects.

10. Detection system of periodic stationary objects according to any of claims.8 and 9, in which the detection module variants periodic stationary objects specifies a predetermined fifth threshold value obtained from the maximum values of the peaks.

11. Detection system of periodic stationary objects according to claim 8 or 9, further comprising a module to determine the frequency, performed with the opportunity to assess whether or not the frequency of the positions where the peaks occur, and lower predetermined fifth threshold value in the evaluation of the interval.

12. The system of detection of� periodic stationary object according to claim 11, in which determining module assesses the frequency, a frequency position in which the peaks occur, and the peaks have values equal to or higher than a predetermined fifth threshold value from the maximum value of the peaks.

13. Detection system of periodic stationary objects according to claim 11, further comprising a detection module lateral movement, is arranged to detect the lateral movement of a moving object, and
when the detection module detects lateral movement lateral movement by a specified distance or more, and lowered a predetermined fifth threshold value, the module determine the frequency of low initializes the fifth threshold value.

14. Detection system of periodic stationary objects according to claim 8 or 9, in which:
module assessment periodic stationary objects:
- saves the position in which the peaks occur, and their peak values, when evaluated that there is a periodicity in the previous processing,
- evaluates what options are periodic stationary objects are periodic stationary objects, in case of detection of a peak having a peak value equal to or higher than a predetermined fifth threshold Zn�means, different from peaks having estimated the periodicity in the current processing, when the average value of peak values having estimated the periodicity in the current processing is not less than the average peak value having estimated the frequency in the previous processing, to a predetermined value or more, and
- estimates that moving objects are present, when the average value of peak values having estimated the periodicity in the current processing, less than the average peak values having estimated the frequency in the previous processing, to a predetermined value or more.

15. Method of detection periodic stationary objects to detect periodic stationary objects in the vicinity of a moving object, containing:
- the stage of image capture, which captures an image of surroundings of a moving object using the device for capturing images mounted on a moving object;
- phase transformation point of view, which perform the conversion of the viewpoint for the image captured by the capture device images to create a picture of the view from the height of bird's flight;
- a step of extracting feature points, which are extracted characteristic�th point on a three-dimensional object from image data in a predetermined area of the image with the bird's eye view for each of the plurality of subareas, included in the predetermined region;
- the step of calculating the data waveform, on which complete data waveform corresponding to the distribution of feature points extracted at the stage of extraction of the feature points in a predetermined region of the image with the bird's eye view;
- a step of detecting information of the peaks, which detect information peaks form data signal;
- a step of detecting variants of the periodic stationary objects on which to assess whether or not a three-dimensional object having the feature point extracted at the stage of extraction of the characteristic points, a variant of the periodic stationary object, on the basis that equals or exceeds either no information peaks predetermined first threshold value, and
- the evaluation phase of the periodic stationary objects, which define what version of the periodic stationary object is a periodic stationary object when the option of periodic stationary object is detected in the detection step variants of the periodic stationary objects, and the detection is carried out when the predetermined condition.



 

Same patents:

FIELD: transport.

SUBSTANCE: invention relates to technique of driving condition representation to driver. Information representation device to be used in a vehicle includes movement conditions receiving module; the first module for movement conditions distributions calculation; the first module for driving instability determination; module for training completion determination; the second module for driving instability determination; instability selection module and module for information representation to driver. Driving conditions data is the data received in time range differing from immediately preceding time range. Other time range changes depending on training degree which is ratio of elapsed time from movement conditions data collection start to present time to training time, where time range is set so that it becomes greater while training degree increases. In another version, the device additionally contains module of driving scenario determination.

EFFECT: functionality enhancement.

17 cl, 16 dwg

FIELD: transport.

SUBSTANCE: proposed device comprises image capture device mounted at the vehicle to catch the imaged sideways therefrom, nearby vehicle detection unit light projecting candidate object detector projecting the light with luminescence equal to or higher than the first threshold magnitude, unit to detect the causes for revealing if light projecting candidate object exists or not as well as supplier of data to driver on nearby vehicle.

EFFECT: higher accuracy of nearby vehicle.

9 cl, 9 dwg

FIELD: physics, navigation.

SUBSTANCE: invention is intended to prevent collision of vehicles. When monitoring the surrounding space, a first initial state is obtained for a first object (18) located in the surrounding space (14) as a reference point during a first stopping event, wherein the initial state includes an indication of the initial position of the first object (18). During a second stopping event (S1), a determination (S2) is made of whether a vehicle (10) has moved between the first and second stopping events by a given minimum distance and/or whether a given minimum time has passed. The initial state is refreshed (S3) by indicating the position of at least a second object (18) located in the surrounding space (14) as a reference point if it is determined that a vehicle (10) has moved between the first and second stopping events by a distance exceeding the given minimum value and/or that a time greater than the given minimum time has passed. The actual position of the second object (18) is determined (S5) as the actual state for further intent to begin movement. A possible collision situation is detected (S6) if comparison of the actual and the initial state yields a distance greater than a given value. The invention also relates to an anti-collision apparatus.

EFFECT: warning a driver on presence of objects in the surrounding space.

8 cl, 2 dwg

FIELD: transport.

SUBSTANCE: in the first version, the method includes determination of presence of a vehicle moving ahead along adjacent lane in concurrent direction using video camera, measurement of distance to the vehicle with calculation of its side mirrors "blind area" position and rear vehicle position relative to "blind area", as well as determination of its type and dimensions, warning about hazardous proximity using indicators located in instrument cluster, assistance in the process of vehicle driving through vehicle deceleration in case when driver of vehicle moving ahead along adjacent lane begins to execute unsafe manoeuvre. In the second version, monitoring system switches on video camera, ultrasonic sensors and microcontroller. The ultrasonic sensors located on vehicle front bumper determine position of moving ahead vehicle side mirrors "blind area". Microcontroller produces warning signals for driver about vehicle entry into "blind area" of going ahead vehicle.

EFFECT: higher traffic safety due to automatic monitoring of moving ahead vehicle blind area condition.

7 cl, 3 dwg

FIELD: transport.

SUBSTANCE: alert system comprises annunciator of approach to dangerous point connected via communication line with receiver arranged at locomotive, decoder, memory unit, gate multivibrator, three-input OR element, control circuit, reset pulse generator and people presence control pickups.

EFFECT: higher safety.

1 dwg, 1 tbl

FIELD: transport.

SUBSTANCE: invention relates to automotive parking helps. Control device comprises computer to determine the distance to parking point in parking lot for vehicle to overcome, camera to register situation nearby the vehicle, interface for connection to device to output instructions on manoeuvring at motion to parking point, interface for connection with device for manual input of parking space boundaries and interface for connection to means intended for setting relationship between parking space boundaries and vehicle position. Camera is set to view the space ahead of vehicle. Portable computer comprises above described control device. Proposed method comprises registration of situation nearby vehicle by said camera, calculation of distance to parking point in parking lot for vehicle to overcome, outputting of instructions on manoeuvring at motion to parking point Picture produced by camera is used for manual input of parking space boundary to be compared with current position of vehicle.

EFFECT: efficient parking help for driver.

11 cl, 4 dwg

FIELD: transport.

SUBSTANCE: set of inventions relates to device and method for parking assistance. Device contains available parking space recorder, final target parking position setting facility, contact determination facility, final target parking position calculation facility, the first, the second and the third path calculation facility, the first path, the second path and the third path display facility. The contact determination facility provides determination whether there is the first path for vehicle to reach final target parking position without contacting with the edge of available parking space. The method consists in the following: available parking space is recorded, final target parking position is set, existence of the first path for vehicle to reach final target parking position without contacting with the edge of available parking space is determined. The first path is displayed if it exists. Angle of approach and not final target parking position is calculated if the first path does not exist. Not final target parking position is position that should be reached when vehicle is entering available parking space at approach angle. The second and the third paths are calculated and displayed.

EFFECT: higher efficiency of parking assistance.

10 cl, 15 dwg

FIELD: physics.

SUBSTANCE: device according to the first version includes a steering angle detection unit, an operation detection unit in which the steering wheel returns to a neutral position after turning right or left, and a parking mode selection unit. The device according to the second version includes a steering angle detection unit, an operation detection means in which the steering wheel returns to a neutral position after turning right or left, and a parking mode selection means. The method includes detecting the steering angle of the steering wheel, detecting an operation in which the steering wheel returns to a neutral position after turning right or left, selecting one parking mode on the right side of the vehicle or left side of the vehicle from a plurality of parking modes.

EFFECT: easy establishment of the desired parking position.

21 cl, 29 dwg

FIELD: transport.

SUBSTANCE: invention relates to aviation, in particular to methods of navigation assistance for aircraft path determination. Navigation assistance method consists in determination of future approach path using evaluation of forecasted safe radiuses on future path based on calculation of limit moment starting from which a forecasted safe radius exceeds or is equal to alarm signal generation limit and calculation of limit departure moment which corresponds to maximum moment when an aircraft should leave its predetermined path along which it has been moving, to be able to go to safe altitude.

EFFECT: limitation of departure procedures utilisation during loss of satellite navigational information thus reducing airspace saturation and limiting costs and duration of flights.

15 cl, 5 dwg

FIELD: transport.

SUBSTANCE: invention relates to vehicle driving control device. The device contains side object detector, control device to actuate lateral movement inhibition control, start of entering adjacent lane detector, actuation suppressor to suppress actuation of lateral movement inhibition control. Lateral movement is followed by changing vehicle transversal position relative to lane. Suppression of actuation of lateral movement inhibition control is performed when a vehicle has started to enter adjacent traffic lane, and side object is detected in the adjacent traffic lane.

EFFECT: prevention of improper interference in control for vehicle lateral movement inhibition during changing traffic lane.

22 cl, 11 dwg

FIELD: physics.

SUBSTANCE: method comprises image correction and unweighted full image segmentation (SEG) in snippets at the image preprocessing step (FPP), the use of a threshold value algorithm based on the detected local contrast changes at the detection step (DET), monitored learning with the aid of the support vector method (SVM) with a hyperplane (HE) for subdivision into two classes at the classification step (CLA), automatic calculation of the distance of the detected thermal signature of the marine mammal (TSMM) and its variation over time and space relative to the ship (RV) at the localisation step (LOC), the possibility of an instantaneous check of the decision by an operator at the verification step (VER), and automatic provision of user-specific image data (IRV) of the detected thermal signatures of marine mammals (TSMM) at the documentation step (DOC). An infrared camera system (IRC) with full or partial surround monitoring and an active gyroscopic stabilisation relative to the horizon (HZ) are used to generate data.

EFFECT: higher accuracy of real-time capture of marine mammals.

21 cl, 9 dwg

FIELD: physics.

SUBSTANCE: method comprises, at the sender, dividing an electronic image into macroblocks, each macroblock divided into N≥2 blocks from which K<N synchronisation blocks are selected; calculating a transmission synchronisation sequence and embedding into synchronisation blocks which allow embedding; at the recipient, establishing synchronisation of the digital watermark in the received electronic image for which, beginning with the selected initial point of the recipient, the received electronic image is successively divided into macroblocks and blocks, from which blocks for the intended synchronisation are selected; extracting therefrom verification subsequences and merging into a verification sequence which is compared bitwise with all shifts of the calculated receiving synchronisation sequence; the electronic image received by the recipient is considered an electronic image with established synchronisation of the digital watermark, corresponding to the intended synchronisation sequence with the least number of mismatches. The disclosed method can be used to increase the probability of establishing synchronisation of a digital watermark of an electronic image, divided into component parts of an arbitrary size, and for preventing visually noticeable distortions caused by embedding into blocks of the electronic image with virtually unchanged statistical characteristics of transmission synchronisation sequences.

EFFECT: improved establishment of synchronisation of a digital watermark of an electronic image when dividing the electronic image with an embedded digital watermark into component parts of an arbitrary size.

4 cl, 9 dwg

FIELD: information technologies.

SUBSTANCE: method is based on monitoring and input of signals with subsequent computer processing, identification of dynamic individual criteria of motor responses and current psychophysiological conditions of identified persons, generation of reference motor responses of identified persons, comparison with information produced during the first and further contacts with identified persons, for this purpose the text is reproduced on a graphic tablet, functions on its output are scaled by their reduction to standard duration and intensity, spaces are formed with criteria of identified persons by features of their signatures, identified persons are monitored through registration of physiological parameter reflecting their psychophysiological condition, the range of physiological parameter values is divided into subranges, a password is composed, assigned to each identified person. During comparison of criteria they detect numbers of the reference subrange, for which similarity of compared descriptions is maximum, and by the number of this subrange the decision is made on psychophysiological condition of the identified person, and by the reference number the decision is made about the identified person as they are.

EFFECT: increased reliability of personal identification when writing a password.

4 cl, 7 dwg

FIELD: physics.

SUBSTANCE: device further includes an OR logic element and a signature coefficient memory unit, address inputs of which are connected to outputs a shift register unit. All inputs of the shift registers are connected to outputs of the shift register unit, and the reset inputs of all shift registers are integrated and connected to the reset input of the shift register unit; outputs of the signature coefficient memory unit are connected to outputs of the device and to inputs of the OR logic element, the output of which is connected to the second control input of a control unit, the fifth and sixth outputs of which are respectively connected to the control input for accessing the signature coefficient memory unit and to the reset inputs of the shift register unit and a most significant bit address counter.

EFFECT: high efficiency of image recognition.

9 dwg

Vision system // 2538336

FIELD: electricity.

SUBSTANCE: invention is referred to the area of images receiving and processing. In the system the signal is delivered through light filters to the surface of cells filled with a substance polarised under action of the controller command for its light flux control. Upon passage through the light filter the light flux diffracts to slots and passes through transparent medium deviating per a certain angle equal to the wave length and coming to a photocell being a control element in the electric circuit, which allows current passing through a branch with the control element by increase or decrease of current or voltage. Identification and survey control are carried out with holographic techniques of identification by forming displays out of semiconductor elements using data on parameters of the sounding signal and information on environmental conditions from the built-in database.

EFFECT: provision of beam control inertia-free system with electroscopic regulating circuit of the signal parameters.

4 dwg

FIELD: physics, computer engineering.

SUBSTANCE: invention relates to means of recognising objects on an image. The method includes recording the first and second longitudinal profiles of a target area, recorded after a given time interval with respect to each other; establishing a height difference profile based on the said longitudinal profiles; dividing the height difference profile into equidistant horizontal sectional planes of the height; calculating positions of centres of gravity of the areas surrounded by corresponding horizontal lines of the horizontal sectional planes of the height of the areas; sending the established height difference profile and the calculated centres of gravity of areas to an object classification system.

EFFECT: high accuracy of recognition owing to the correlation of horizontal lines on an image.

5 cl

FIELD: physics, signalling.

SUBSTANCE: invention relates to labelling means for identification of articles. In the method, an assigned unique number is converted to a two-dimensional bar code of a Data Matrix symbol. At least two said bar codes are placed on the surface of an article in the immediate vicinity of each other. The distance between said bar codes is random (arbitrary). The obtained label consists of a digital part which identifies the article and an analogue part which is defined by the mutual arrangement of the bar codes and which authenticates the article. When simultaneously reading said bar codes, coordinates of their corners in one coordinate system are calculated. An article is authenticated by simultaneously reading the bar codes and comparing with information contained in a data base.

EFFECT: safer labelling with authentication.

5 dwg

FIELD: physics, computer engineering.

SUBSTANCE: invention relates to means of accessing a processing system. The method of accessing a processing system comprises steps of: receiving an ordered sequence of biometric data inputs from a user through a biometric sensor connected to the processing system; determining presence of a match of each of the biometric data inputs of the ordered sequence with a corresponding record stored in the table of non-volatile memory of the processing system, wherein the table includes the stored ordered sequence of biometric data inputs corresponding to a user password combination; and if there is match, providing the user with access to the processing system, and if there is no match, denying the user access to the processing system, wherein each of the biometric data inputs of the ordered sequence corresponds to a different finger of the user and an element of the password combination selected by the user, wherein the first finger of the user corresponds to the first alphanumeric character selected by the user, and the second finger of the user corresponds to the second alphanumeric character selected by the user.

EFFECT: high reliability of access to a processing system.

20 cl, 6 dwg

FIELD: radio engineering, communication.

SUBSTANCE: invention relates to radio signal discrimination methods, in particular, to the methods of discriminating the radio signal modulation type and parameters. The invention can be used in devices used for the discrimination of radio signals of a complex frequency and timing structure. The method involves the preliminary generation of energy distribution matrices from discretised and quantised counts of reference radio signals (RS), based on their frame wavelet conversions. Next, marker vectors are formed from these matrices by the line-by-line concatenation of all wavelet coefficients. After that, the marker vector elements are classified, and each of RS used for the identification is divided into fragments. Duration of every fragment is equal to the one of a reference RS. Then, the average value fragment for RS under discrimination is calculated. This fragment is then used for its marker vector generation, in the same way as for the reference RS. Then the discriminated RS is identified by subtracting the absolute value of its marker vector from the absolute values of the marker vectors for each of the reference RS. The discriminated RS is incidental to the reference RS with the least difference between the marker vectors.

EFFECT: higher chance of correct radio signal discrimination due to the lower influence to the marker vector formed from the primary radio signal, which is defined by the information component.

5 dwg

FIELD: electricity.

SUBSTANCE: invention is related to a modular electric device comprising a body with the front cover with available identification area, a cover protecting the label, which interacts with the above identification area forming a socket with a groove open outside and a tab, which is setup so that it may be moved in the above socket through the groove. At that the above tab is equipped with edge along one side thus it covers the above groove, at least partially, when the tab is set up in the socket completely. In compliance with invention the modular electric device contains a fixing unit preventing movement of the tab in regard to the body when the tab is in retracted position outside the socket. At that the tab is made as a plate with the front side surrounded by a fin, which forms cavity for the label, where it has local decrease in height.

EFFECT: the suggested invention allows prevention of the tab movement in regard to the body when it is outside the socket thus ensuring free access to the front side of the above tab in order to place the identification label of the modular device in the position or to remove it.

15 cl, 9 dwg

FIELD: transport.

SUBSTANCE: invention relates to technique of driving condition representation to driver. Information representation device to be used in a vehicle includes movement conditions receiving module; the first module for movement conditions distributions calculation; the first module for driving instability determination; module for training completion determination; the second module for driving instability determination; instability selection module and module for information representation to driver. Driving conditions data is the data received in time range differing from immediately preceding time range. Other time range changes depending on training degree which is ratio of elapsed time from movement conditions data collection start to present time to training time, where time range is set so that it becomes greater while training degree increases. In another version, the device additionally contains module of driving scenario determination.

EFFECT: functionality enhancement.

17 cl, 16 dwg

Up!