# Method and device of video processing, and recording medium that stores program

FIELD: information technologies.

SUBSTANCE: target image that forms video image is divided into multiple division areas (DA); pass band (PB) width applied to DA is determined; array of filtration ratios (FR) is calculated to realise frequency characteristics corresponding to limitation of band, with application of PB width; image data is filtered with application of FR array; error information value is produced between obtained data and data of initial image, and distribution ratio (DR) is calculated to be used to determine optimal width of PB, on the basis of produced value; optimal width of PB corresponding to DR is defined for each DA, and array of optimal FR is calculated to realise frequency characteristics corresponding to limitation of band, using optimal width of PB; image data of division area is filtered using array of optimal FR; and produced data of each DA are synthesised.

EFFECT: generation of filtered image with specified value of image quality assessment.

29 cl, 27 dwg

The technical FIELD TO WHICH the INVENTION RELATES.

The present invention relates to a video processing method and a corresponding device used to perform the operation of the simplified filter, which adaptively applies to the images forming the video image, and also relates to a video processing program used for implementing the method of video processing, and machine-readable storage medium that stores the program.

This application claims the priority, application Japan No. 2006-353610, filed December 28, 2006, the contents of which are incorporated herein by reference.

The LEVEL of TECHNOLOGY

It is known that the pre-filter, which is often used during preview processing for encoding video information, is effective for reducing block distortion, distortion in the form of points on the edges of objects ("mosquito noise"), etc. followed by the encoding, and thus improves the subjective quality of the image. Bandwidth (hereinafter referred to as "bandwidth") used pre-filter is limited in such a way as to reduce the noise contained in the original image, and improve the coding efficiency. However, if bandwidth is narrowed too much, the image quality is extremely hudaida.

Fig illustrates a method of image processing comprising the limitation of bandwidth.

As shown in Fig, in the method of image processing comprising the limitation of bandwidth, first data B(1) of the original image is entered and then converted into a frequency component of I(1) (see step S1000). Frequency component of I(1) subject to the limitation of bandwidth using bandwidth r1 (0<r1<1) in order to obtain the frequency component of I(r1) (see step S1100). Frequency component of I(r1) is subjected to image conversion, and thus are formed the data B(r1) filtered image (see step S1200).

When such image processing is applied to all frames of the video image through the use of the same bandwidth, the image quality of each of the filtered frame is not the same, because each frame has an individual frequency characteristics of the image. Therefore, an image having a large number of low frequency components is only a small difference from the original image, and thereby the deterioration of the subjective and objective image quality is small. However, in the image having many high-frequency components, edges, etc. are smoothed and blurred, which greatly degrades with bjective and objective quality of the image.

As the objective value of the image evaluation is often used, for example, the peak signal-to-noise ratio (PSNR). When the specified level (S) of the signal and the level (N) noise ratio PSNR is defined by the following formula:

PSNR = 20×log_{10}(S/N)

In actual processing, if the brightness of the original image represented by 8 bits (i.e., from 0 to 255), the ratio PSNR can be calculated using the following formula:

where N denotes the number of pixels of the original image and the filtered image; f(x,y) denotes the value of each pixel of the original image; and f'(x,y) denotes the value of each pixel of the filtered image. In addition, "255" indicates the maximum amplitude (or the pixel value) of each pixel of both images.

Thus, when the actual processing of the original image and the filtered image are compared with each other (namely, using the above formula) for calculating the ratio PSNR.

In the method of solving the above objectives management objective and subjective image quality is performed by "cyclic" restrictions strips applied to each image.

Fig illustrates the device structure 1000 of formation of optimum filtered image is agenia to generate the data of the optimum filtered image by performing the "circular" limitations strip.

As shown in Fig, the device 1000 of formation of optimum filtered image includes the 1100 block of input original image data, the 1200 block of frequency component, block 1300 manual selection of the bandwidth, block 1400 restrictions stripe, 1500 block forming the image data, block 1600 calculating the ratio PSNR, the block 1700 evaluation image and the 1800 block of output data of the image with the optimal limitation of bandwidth.

Fig illustrates a method of processing image data generating optimum filtered image by performing the "circular" limitations strip, and the method is performed in the device 1000 of formation of optimum filtered image having the structure above.

The device 1000 of formation of optimum filtered image data B(1) of the original image is first introduced in the 1100 block of input original image data and then converted into a frequency component of I(1) in the 1200 block of frequency component (see step S2000).

Then in block 1300 manual selection of the bandwidth manually pre-selected width r1 of the strip (see step S2100). Then in block 1400 restrictions band converted frequency component of I(1) subject to the limitation of bandwidth using you the early width r1 of the strip to obtain a frequency component of I(r1) (see step S2200).

Further, in the 1500 block of forming image data of the frequency component (r1) is subjected to image conversion, and thus are formed the data B(r1) of the image (see step S2300). In block 1600 calculating the ratio PSNR data B(1) of the original image are compared with the data B(r1) image for calculating the ratio PSNR (r1) (also called "P(r1)") (see step S2400).

In the 1700 block image evaluation is determined whether the calculated ratio P(r1) the desired quality of the image (see step S2500). If it has the desired image quality, the 1800 block of output data of the image with the optimal limitation of bandwidth issues data B(r1) as the image data with the optimum limitation of bandwidth (i.e. data optimum filtered image) (see step S2600).

However, it happens very rarely that the ratio P(r1)obtained at the first processing step, had the desired image quality. When it is not desired image quality, the processing returns to the process (step S2100), performed by block 1300 manual selection of the bandwidth, and the width (r2) stripes once again chosen so that the corresponding image limited band had a quality that is closer to the desired image quality. Then the limitation of bandwidth, image formation and vychislyayutsya PSNR again performed in the same way.

Thus, the above-described processing is repeated N times until the desired image quality, and the width of rN strip, which is finally obtained, is used as the optimal bandwidth for the data B(rN) image using the 1800 block of output data of the image with the optimal limitation of bandwidth. The generated data B(rN) images are output as the image data with the optimal limited bandwidth (i.e. data optimum filtered image) (see step S2600).

However, in the above-described method, various video and all frames that are generated from them, are filtered, evaluated by subjective and objective image quality of each received image signal, and the corresponding processing is repeated "loop"until it receives an equal image quality for all video frames. Taking into account the required time and cost for processing a large number of images, the above-described method is inappropriate and not suitable for use.

In order to solve the above problem, a well-known technique (see patent document 1) image processing is performed using the optimal bandwidth based on the data encoding (video) the picture is of.

Fig illustrates the structure of the device 2000 of formation of optimum filtered image data generating optimum filtered image using the data encoding.

As shown in Fig, the device 2000 of formation of optimum filtered image includes the 2100 block of input original image data, the 2200 block of frequency component, of the 2300 block coding of the image data, block 2400 determine the optimal limited bandwidth, block 2500 limitations strip, block 2600 forming the image data and the 2700 block of output data of the image with the optimal limitation of bandwidth.

Fig illustrates a method of processing image data generating optimum filtered image through the use of data encoding, and the method is performed in the device 2000 of formation of optimum filtered image having the structure above.

In the device 2000 of the formation of the optimal first filtered image data B(1) of the original image is entered in the 2100 block of input data of the original image 2100 and then converted into a frequency component of I(1) in block 220 analysis of the frequency component (see step S3000).

Then, in the 2300 block coding of the image data in adnie data B(1) of the original image is encoded (see step S3100). Based on the information about the amount of code obtained by using the appropriate coding, the optimal width of the strip is determined by r1 in the 2400 block of determining the optimal bandwidth (see step S3200).

In block 2500 restrictions band converted frequency component of I(1) subject to the limitation of bandwidth by using a certain width r1 of the strip to obtain a frequency component of I(r1) (see step S3300). In the 2600 block of forming image data of the frequency component of I(r1) is subjected to image conversion, and thus are formed the data B(r1) of the image (see step S3400).

Finally, the data B(r1) images are output as the image data with the optimum limitation of bandwidth (i.e. data optimum filtered image) from the 2700 block of output data of the image with the optimal restriction of the strip (see step S3500).

Accordingly, in the conventional device 2000 of formation of optimum filtered image, made, as shown in Fig, after performing the encoding is determined by the optimal bandwidth based on the data encoding obtained by coding. Therefore, the data of the optimum filtered image obtained without performing repetitive processing, as required in the device 1000 to build the project for optimum filtered image, performed, as shown in Fig.

Patent document 1: unexamined application for patent of Japan, first publication No. H06-225276.

__DISCLOSURE of INVENTION__

**The problem solved by the invention**

Of course, in accordance with the traditional device 2000 of formation of optimum filtered image, made, as shown in Fig, the data of the optimum filtered image can be formed without performing repetitive processing, as required in the device 1000 of formation of optimum filtered image, made, as shown in Fig.

However, in the device 2000 of formation of optimum filtered image shown in Fig, after performing encoding the optimal bandwidth is determined on the basis of information coding, obtained through the encoding.

In this way, using the data coding process limitations strip and the encoding process are inseparable. Therefore, even if the user would like to perform pre-filtering process using the optimum bandwidth, coding is also required. If the encoding is also performed after pre-filtering, coding will be performed twice. In particular, if the size of the image is great it will take considerable processing time.

Taking into account the above, in order to optimize bandwidth for pre-filter, it is preferable to use a method that can facilitate appropriate treatment and may intentionally be managed using a standard for assessing the subjective or objective image quality, compared with the method using data encoding (for example, the amount of code).

In light of the above circumstances, the present invention is to provide a new method of image processing, by means of which can be implemented adaptive filtering process for images that form a video without encoding process and without repeating the processing with regard to the frequency distribution in the frame or between frames of images, and thereby effectively formed a filtered image that has the specified value assessment of the quality of the image.

**Means for solving the problem**

A: the First structure

To solve the above-mentioned task, the video processing device according to the present invention includes: (1) a separation unit for separating the target image, which forms a video image into many areas of the division; (2) unit definition per the th bandwidth to the first bandwidth, applied to the areas of separation, separated by the separation unit; (3) the computing unit of the first array of filter coefficients to compute a first array of filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, by using the first bandwidth determined by the block determining the first bandwidth; (4) the shaping unit data filtered divided image data generating filtered divided image for each area of separation (separated by the separation unit through the exposure of the image data of each area of separation (separated by the separation unit) to the filtering process using the first array of filter coefficients calculated by block calculating a first array of filter coefficients; (5) the block coefficient calculation distribution to derive for each area of the separation values for the error information between the image data of each area separation data and the filtered divided image formed by the forming unit data filtered divided image, and calculate the distribution factor used to determine the protected area the maximum bandwidth, on the basis of the extracted values; (6) block determine the optimal bandwidth for determining for each region, division, separated by the separation unit, the optimum bandwidth corresponding to the distribution coefficient calculated by the computing unit distribution ratio; (7) the computing unit of the array of optimal filter coefficients to calculate for each region, division, separated by the separation unit, the array of optimal filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, using the optimum bandwidth determined by the block determining the optimal bandwidth; (8) the block data generating optimum filtered divided image data generating optimum filtered divided image of each area of separation (separated by the separation unit through the exposure of the image data of each region separating the filtering process using an array of optimal filter coefficients calculated by the computing unit of the array of optimal filter coefficients; and (9) block synthesis for the synthesis of optimal data otherthrow the frame is divided image of each divided region, which were formed by the block data generating optimum filtered divided image.

The above-described structure may further include:

a Comparer for comparing for each area of the separation optimal bandwidth determined by the block determining the optimal bandwidth, with an optimal bandwidth of the peripheral area of separation around this area separation; and

the power adjustment to adjust the optimal bandwidth determined by the block determining an optimum bandwidth, based on the comparison result.

The above-described structure may further include:

the block definition to determine whether the moving image data of each region, division, using the image data of the frame before or after the frame area or image data frames before and after the frame region; and

the power adjustment to adjust the optimal bandwidth (determined by the block determining the optimum bandwidth) of each area of separation, for which by definition block is determined that the image data area splitting have movement.

With what you learn may that:

the definition block determines whether the moving image data area splitting and are characterized by whether they are high-frequency component; and

block adjustment of the optimal bandwidth adjusts certain optimal bandwidth each area of separation, which determined that these images have movement and are characterized by high-frequency component.

The rendering of the present invention, which is implemented when the above-described units are operated, can also be implemented through computer program. Such a computer program, the machine can be provided through a store on the corresponding machine-readable storage medium or via the network and can be installed and run on the control unit, such as a Central processor in order to implement the present invention.

B: the Second structure

To solve the above-mentioned tasks another device processing the video data according to the present invention includes: (1) unit determining the first bandwidth to the first bandwidth applied to the size of the area of separation for separation, which are set on the target image processing, forming the video image so that the button be divided into sections destination image processing; (2) the computing unit of the first array of filter coefficients to compute a first array of filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, by using the first bandwidth determined by the block determining the first bandwidth; (3) the set of filtered image data to generate the data of the filtered image by the exposure data of the target image processing filtering process using the first array of filter coefficients calculated by the computing unit of the first array of coefficients of the filter; (4) the block coefficient calculation distribution to derive for each area of the separation values of the errors between the data of the target image processing and data filtered image generated by the processing unit data of the filtered image, and calculate the distribution coefficient used for determining an optimum bandwidth, based on the extracted values; (5) the unit determine the optimal bandwidth for determining for each area of the separating optimum bandwidth corresponding to the distribution coefficient, Vici is certain by means of the computing unit distribution ratio; (6) the computing unit of the array of optimal filter coefficients to calculate for each area of the separation of the array of optimal filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, using the optimum bandwidth determined by the block determining the optimal bandwidth; (7) the block data generating optimum filtered divided image data generating optimum filtered divided image of each region dividing by the exposure of the image data of each region separating the filtering process using an array of optimal filter coefficients calculated by the computing unit of the array of optimal filter coefficients; and (8) block synthesis for the synthesis of optimal data filtered split each image region separation, which were formed by the block data generating optimum filtered divided image.

The above-described structure may further include:

a Comparer for comparing for each area of the separation optimal bandwidth determined by the block determining the optimal width of polypropylene, with the optimal bandwidth of the peripheral area of separation around this area separation; and

the power adjustment to adjust the optimal bandwidth determined by the block determining an optimum bandwidth, based on the comparison result.

The above-described structure may further include:

the block definition to determine whether the moving image data of each region, division, using the image data of the frame before or after the frame area or image data frames before and after the frame region; and

the power adjustment to adjust the optimal bandwidth (determined by the block determining the optimum bandwidth) of each area of separation, for which by definition block is determined that the image data area splitting have movement.

In this case, it is possible that:

the definition block determines whether the moving image data area splitting and are characterized by whether they are high-frequency component; and

block adjustment of the optimal bandwidth adjusts certain optimal bandwidth each area of the division for which it is determined that its image data have movement and are characterized by high-frequency component.

The rendering of the present invention, which is implemented when the above-described units are operated, can also be implemented through computer program. Such a computer program, the machine can be provided through a store on the corresponding machine-readable storage medium or via the network and can be installed and run on the control unit, such as a Central processor in order to implement the present invention.

C: the Processes of the present invention

In the video processing device having the first structure of the present invention, when the target image processing, forming the video image, it is divided into many regions of separation. The first bandwidth applied to the areas of separation, is determined, for example, based on the size of each field division.

Then, it calculates the first array of filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, by using the first bandwidth, and to derive the filtered divided image of each region dividing by the exposure of the image data of each region separating the filtering process using the calculated first array of filter coefficients./p>

Then for each area of the division is displayed (for example, the ratio PSNR) information about the error between the image data of each field division and generated data filtered divided image, and the distribution coefficient used for determining an optimum bandwidth, is calculated on the basis of the extracted values.

For example, we compute the distribution factor by dividing the value of the error information, which is received in a state very close to a state in which there is limitation of bandwidth, the displayed value of the error information.

On the other hand, in the video processing device having the second structure of the present invention, when the target image processing, forming the video image may be determined first bandwidth based on the size of the area of separation for separation, which are set on the target image processing, so as to be divided into sections destination image processing, and thus is determined by the first bandwidth applied to the areas of separation.

Then, it calculates the first array of filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, by using the first sirinyali bandwidth and to derive a filtered image by the exposure data of the target image processing filtering process using the calculated first array of coefficients of the filter.

Then for each area of the division is displayed (for example, the ratio PSNR) information about the error between the target image and processing the generated data of the filtered image, and the distribution coefficient used for determining an optimum bandwidth, is calculated on the basis of the extracted values.

For example, we compute the distribution factor by dividing the value of the error information, which is received in a state very close to a state in which there is limitation of bandwidth, the displayed value of the error information.

After the calculated distribution coefficient for each area of the division, as described above, in the first and second structures executes the same processing.

Thus, then for each area of the division is determined optimum bandwidth corresponding to the calculated distribution coefficient, for example, by reference to the table determine the optimal bandwidth, which is configured to fit between the ratios is ntom distribution and the optimal bandwidth.

When given multiple tables determine the optimal bandwidth in accordance with the size of the image and the target value of the error information, select table determine the optimal bandwidth, which corresponds to the size of the area of separation and a given target value of the error information, and the optimum bandwidth corresponding to the distribution coefficient, determined by reference to the selected table to determine the optimal bandwidth.

Then for each area of the division calculates the optimal array of filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, using a specific optimal bandwidth, and to derive optimum filtered divided image of each region dividing by the exposure of the image data of each region separating the filtering process using the calculated optimal array of coefficients of the filter.

In the last step, the data of the optimum filtered divided image are synthesized, and thereby, the filtered image to the destination image processing.

As described you the e invention the filtering process to convert the target image processing in the image, having fixed the value of assessing the quality of image can be performed automatically without coding and without repetitive processing.

In the present invention having the structure described above, each area of separation is subjected to a filtering process using an array of optimal filter coefficients computed for the area of separation. Therefore, the resulting filtered image is generated for the target image processing may include noise at the boundaries of the fields.

Therefore, in the example, the optimal bandwidth defined for each area of the division is compared with the optimal bandwidth of the peripheral area of separation around this area of separation, and if between them there is a great difference, a certain optimal bandwidth is adjusted so as to reduce the difference.

For optimal bandwidth each area of separation defined in the present invention, when the area division is part of the image, in which there is movement, even if the optimal bandwidth is reduced (which may reduce the amount of code), the image data area splitting can continue to have the same subjective image quality compared to other areas of the division, although their objective image quality is not equal to the objective image quality of other areas in the division.

Taking into account the above possible:

to determine whether the moving image data of each region, division, using the image data of the frame before or after the frame area or image data frames before and after the frame region (for example, by estimating variations of the pixel values for that frame and the current frame); and

to adjust certain optimal bandwidth each area of the division for which it is determined that the image data area splitting have movement, in order to reduce the optimal bandwidth.

For optimal bandwidth each area of separation defined in the present invention, when the area division is part of the image that has movement and is characterized by high-frequency component, even if the optimal bandwidth is greatly reduced (which can greatly reduce the amount of code), the image data area splitting can continue to have the same subjective image quality compared with other areas of the division, although their objective image quality is not equal to the objective quality of the image is Oia other areas of the division.

Given the above possible:

to determine whether the image data of each region of the separation movement and are characterized by whether they are high-frequency component, using the image data of the frame before or after the frame region, or the image data frames before and after the frame region (for example, by definition, illustrates whether a value that indicates the attribute of the image data area splitting that image data are characterized by high-frequency component, and at the same time by evaluating the variation of the amount of areas of separation, which show that the corresponding image data are characterized by high-frequency component, for such a frame (used for definition) and the current frame); and

to adjust certain optimal bandwidth each area of the division for which it is determined that the image data area splitting have movement and are characterized by high-frequency component, in order to reduce the optimal bandwidth.

Accordingly, even when the image forming part of the video that is part of the image that includes many high-frequency components, and the part of the image that does not include many high-frequency SOS is alausa, data optimum filtered image for the implementation of the target values of the error information (for example, the target relations PSNR) can be formed for each part of the image.

**The effect of the invention**

In accordance with the present invention the adaptive filtering process for images that form a video image, can be implemented without coding and without repeating the processing with regard to the frequency distribution in the frame or between frames of images, and thereby effectively formed a filtered image that has the specified value assessment of the quality of the image.

__BRIEF DESCRIPTION of DRAWINGS__

1 is a diagram used to explain the results of experiments to obtain the relationship of correspondence between the bandwidth and the ratio PSNR.

2 is a diagram used for explanation of the table determine the optimal bandwidth.

Figure 3 - diagram used for explanation of the table determine the optimal bandwidth.

4 is a diagram showing the structure of the device forming the optimum filtered image as a first variant implementation of the present invention.

Figa diagram used for explanation of the table define the first width is not the bandwidth.

Figw - diagram used for explanation of the table identify the first bandwidth.

6 illustrates a block diagram of the sequence of operations performed by the device forming the optimum filtered image of the first variant implementation.

Fig.7 is a diagram showing the structure of the device forming the optimum filtered image as a second variant implementation of the present invention.

Fig illustrates a block diagram of the sequence of operations performed by the device forming the optimum filtered image of the second variant implementation.

Fig.9 is a diagram showing the structure of the device forming the optimum filtered image as the third alternative implementation of the present invention.

Figure 10 illustrates a block diagram of the sequence of operations performed by the device forming the optimum filtered image of the third variant implementation.

11 is the block diagram of the sequence of operations performed by the device forming the optimum filtered image of the third variant implementation.

Fig diagram for explaining the process of adjustment to the optimal bandwidth, performed by the block korrektirovki the optimal bandwidth.

Fig is a diagram for explaining the process of adjustment to the optimal bandwidth, performed by the block adjustment of the optimal bandwidth.

Fig is a diagram for explaining the process of adjustment to the optimal bandwidth, performed by the block adjustment of the optimal bandwidth.

Fig is a diagram for explaining the process of adjustment to the optimal bandwidth, performed by the block adjustment of the optimal bandwidth.

Fig diagram showing the structure of the device forming the optimum filtered image as a fourth variant of implementation of the present invention.

Fig illustrates a block diagram of the sequence of operations performed by the device forming the optimum filtered image of the fourth version of the implementation.

Fig also illustrates the block diagram of the sequence of operations performed by the device forming the optimum filtered image of the fourth version of the implementation.

Fig diagram showing the structure of the device forming the optimum filtered image as a fifth variant of the implementation of the present invention.

Fig illustrates a block diagram of the sequence is eljnosti operations, performed by the device forming the optimum filtered image fifth variant of the implementation.

Fig diagram for explaining the image in the frame.

Fig diagram showing the results of an experiment performed to confirm the effectiveness of the present invention.

Fig diagram used to explain the method of image processing comprising the limitation of bandwidth.

Fig illustrates the structure of the device forming the optimum filtered image data generating optimum filtered image by performing the "circular" limitations strip.

Fig - sequence of operations performed by the device forming the optimum filtered image, to generate the data of the optimum filtered image by performing a "cyclic" limitations strip.

Fig illustrates the structure of a traditional formation of the optimal filtered image.

Fig - sequence of operations performed by a traditional device for the formation of the optimum filtered image.

**Notation**

1 | the forming device Optim is a high filtered image |

100 | the input block of the original image data |

101 | the separation unit of the image |

102 | unit determining the first bandwidth |

103 | table determining the first bandwidth |

104 | the computing unit of the first array of coefficients of the filter |

105 | the block data generating filtered image |

106 | the computing unit distribution ratio |

107 | the block selection table determine the optimal bandwidth |

108 | table determine the optimal bandwidth |

109 | block determine the optimal bandwidth |

110 | the computing unit of the array of optimal filter coefficients |

111/td> | the block data generating filtered image |

112 | data collection unit |

113 | the synthesis block data of the filtered image |

200 | part of the repetition |

__The IMPLEMENTATION of the INVENTION__

In order to solve the problems described above relating to traditional methods, the authors present invention invented the invention, in which first determine preliminary bandwidth in accordance with the size of the image data for the target image processing, and based on the size of the image form the preliminary data of the filtered image in order to measure the objective value assessment of image quality. Then we can calculate the dimensionless parameter, such as the distribution coefficient based on the measured values of the objective assessment of image quality, and come to the table determine the optimal bandwidth using the calculated distribution coefficient as a key to determine the optimal bandwidth for the implementation of the target values of the evaluation of objective quality images of the I, the table determine the optimal bandwidth has the structure of a data transformation in which the more the distribution coefficient of the image data, the more defined the optimal bandwidth. Based on the optimal bandwidth form optimum filtered image data for data processing purposes.

In accordance with the above-described invention, the filtering process for converting the original image into an image having the specified value of the objective assessment of image quality can be performed automatically without the process of coding and non-recurring operations, in order to solve traditional problems.

In the above invention, one image is a goal for the corresponding filtering process.

However, one image has a part, which includes many high-frequency components, and a portion that does not include many high frequency components. Therefore, when the whole image is subjected to a filtering process using a single array of filter coefficients, the image quality is considerably degraded in parts, which includes many high-frequency components, and the image quality is not so ear which agrees in part, which does not include many high frequency components.

In addition, the above-described invention is provided by taking into account the limitations of the band, using the metric objective picture quality, but without taking into account the limitations of the band, using the metric of subjective image quality. However, the limitation of bandwidth by using not only the metric objective picture quality, but also the indicator of subjective image quality can be preferred with regard to coding efficiency.

That is, for a high-frequency component in an area where there is movement, for example for high-frequency component related to water flow, or fireworks, or for the high-frequency component obtained as a result of rapid panning of the camera, there is a significant visual deterioration compared with the high-frequency component in the region where movement does not occur.

Therefore, to further improve the coding efficiency, it is preferable to perform the filtering process in which a high-frequency component region having motion, and the high-frequency component region having no motion, are assigned to different indicators objective image quality (for the relationship PSNR relative to lower the ratio PSNR is assigned to the high-frequency component area, with movement). Thus, it is preferable to determine the presence or absence of motion for each of the target high-frequency component and to apply adaptive weighting to the high-frequency component to specify the bandwidth and perform the appropriate filtering process.

Taking into account these circumstances, the authors of the present invention invented an additional improvement of the invention described above.

Next will be specifically explained the reason for the filtering process for converting a target image processing in the image that has the specified value assessment of image quality, can be performed automatically without coding and without repeating the operation.

For convenience, subsequent clarifications while maintaining the generality of the explanations separation image is not considered, and the ratio PSNR is used as the error information.

Figure 1 illustrates the results of the experiments for obtaining the correlation between the ratio PSNR of each corresponding image (see "P(r)" in figure 1) and the width r of the strip, with five different images 1-5 were used as images for experiments, and filtering was applied to the data of the image (namely, the brightness components) through the use of an array of coefficie the tov filter for implementing frequency characteristics, the corresponding equal bandwidth r (0,3<r<1) and in the horizontal and vertical directions. Each image has a size of 1920×1080 pixels.

As described above, in the present invention, the first width of the strip is determined by r1 at the first stage. For example, it is assumed that the first width r1 of the strip is set equal to 0.5.

In the next step of the present invention are formed filtered image data using a first array of filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth using r1=0.5 and is calculated ratio PSNR for the data of the filtered image. Thus, when the image processing 1-5, having the characteristics shown in figure 1, to derive a filtered image, and then calculates the value of P(0,5) as a ratio PSNR for such first data from the filtered images.

In accordance with the above calculation, as shown in figure 1, P(0,5)=34,5 for image 1; P(0,5)=42,3 for image 2; P(0,5)=40,6 for image 3; P(0,5)=42,7 for image 4 and P(0,5)=45,3 for image 5.

In the next step of the present invention can be used to calculate the coefficient of X distribution by dividing the value of the relationship PSNR ("51,2" in figure 1), which is obtained in a state very close to the status, in which the limitation of bandwidth is not performed, for each calculated value of the ratio PSNR.

In accordance with the above calculation uses the formula "X=51,2/P(r1), and the distribution coefficient X=1.48 for the image 1; the distribution coefficient X=1,21 for image 2; the distribution coefficient X=1,26 for image 3; the distribution coefficient X=1,20 for image 4; and the distribution coefficient X=1,13 for image 5.

In the next step of the present invention is the process of determining the optimum bandwidth corresponding to the calculated distribution coefficient. Although the identification process can be carried out using the program, namely, the detection may be performed by reference to the table determine the optimal bandwidth, which defined the relationship of correspondence between the distribution coefficient and the optimum bandwidth.

Table determine the optimal bandwidth to which a call is made, can be prepared many tables related to the size of the image and the target ratio PSNR (see Figure 2). Table assigned to the image size and target relations PSNR (see figure 3), manages the information value optimum width r2 strips (use the th for the implementation of target relations PSNR), assigned to each coefficient of X placement within the range of its values.

For example, the relation of correspondence between the range of the X factor host and the optimal width r2 strips (used to implement the target relations PSNR) is controlled so that the optimum width r2 of the strip is equal to: B_{1}for each coefficient of X placement within the range of X<A_{1}; B_{2}for each coefficient of X placement within the range A_{1}≤X<A_{2}; and B_{3}for each coefficient of X placement within the range A_{2}≤X<A_{3}.

The value of A_{i}(i=1,..., n-1) have the following relationship:

0<A_{1}<A_{2}<A_{3}<...<A_{n-2}<A_{n-1}

In accordance with the fact that the larger the coefficient of X distribution, the more optimal the width r2 of the strip, we get the following:

0<B_{1}<B_{2}<B_{3}<...<B_{n-2}<B_{n-1}<B_{n}<1

In accordance with this, the next step of the present invention refer to table determine the optimal bandwidth with the data structure shown in Figure 3, using the calculated coefficient of X host as the key, and thus determines the value of B_{i}as the optimal width r2 of the bands corresponding to value ratios are the NTA X host.

As described above, the table determine the optimal bandwidth has the following table structure:

0<A_{1}<A_{2}<A_{3}<...<A_{n-2}<A_{n-1}

0<B_{1}<B_{2}<B_{3}<...<B_{n-2}<B_{n-1}<B_{n}<1

Thus, the most optimal width r2 band is assigned to the image data having the greater coefficient of the X distribution, and less optimal width r2 band is assigned to image data having a smaller coefficient of the X distribution.

Thus, as is clear from the formula "X=51,2/P(r1)", the image data having the greater coefficient of X distribution have a smaller value of P(0,5); thus, to achieve the target ratio PSNR requires optimal width r2 of the strip. On the contrary, the image data having a smaller coefficient of the X distribution, have a greater value of P(0,5); thus, to achieve the target ratio PSNR, requires less optimal width r2 of the strip.

Considering the above, to show that the most optimal width r2 band is assigned to the image data having the greater coefficient of the X distribution, and less optimal width r2 band is assigned to image data having a smaller coefficient of X distribution table determine the optimal bandwidth is article is ucture table:

0<A_{1}<A_{2}<A_{3}<...<A_{n-2}<A_{n-1}

0<B_{1}<B_{2}<B_{3}<...<B_{n-2}<B_{n-1}<B_{n}<1

The optimal width r2 of the strip, as defined above, is the bandwidth for the data generating optimum filtered image that implement the target ratio PSNR.

In accordance with this, the next step of the present invention calculates the optimal array of filter coefficients for implementing frequency characteristics corresponding to the limits of the band, using the optimum width r2 of the strip, and the corresponding image data are subjected to a filtering process using an array of optimal filter coefficients, i.e. the process adaptive filtering, in which a relatively wide bandwidth is assigned to the image data, which include many high-frequency components, and a relatively narrow bandwidth is assigned to the image data, which do not include many high-frequency components, and thereby to derive the optimum filtered image for the implementation of target relations PSNR.

In accordance with the present invention, the image data should be subjected to only two filtration processes for forming the Oia data, the optimum filtered image for the implementation of target relations PSNR.

Although the above explanation does not consider the separation of the image, the target image processing is divided in the present invention, and each region of the separation described above is subjected to the filtering process, characteristic of the present invention.

Below the present invention will be explained in detail in accordance with alternative embodiments.

(1) the First option exercise

Figure 4 illustrates an example of the structure of the device 1 forming the optimum filtered image as a first variant implementation of the present invention.

As shown in figure 4, the device 1 for generating optimal filtered image as a first variant implementation of the present invention has a block 100 of the input original image data, the block 101 split image unit 102 determining the first bandwidth, the table 103 determining the first bandwidth, block 104 calculation of the first array of coefficients of the filter unit 105 to generate the data of the filtered image, the block 106 calculation of the distribution coefficient, block 107 table selection determine the optimal bandwidth, the table 108 to determine the optimal bandwidth, block 109 determine the optimal bandwidth, block 110 you is ilenia array the optimal filter coefficients, block 111 forming the filtered image data, the block 112 data accumulation and block 113 synthesis data of the filtered image.

The block 105 forming the filtered image data, the block 106 calculation of the distribution coefficient, block 109 determine the optimal bandwidth, the calculation unit 110 of the array of optimal filter coefficients and the block 111 forming the filtered image data to process every data B(1) is divided into image elements generated by the block 101 split image, and, thus, form part of the 200 repetitions.

Block 100 input original image data enters the device data B(1)_all the original image that are to process and form the image.

Based on the size D of the element or the number of E parts division, which are determined in advance, the unit 101 split image divides the image data entered through block 100 input original image data, to generate the data B(1) is divided into image elements of the original image. Although the shape of each element is not limited, for convenience of further explanation assumes that the rectangle.

Block 102 determining the first bandwidth refers to table 103 definitions per the th bandwidth, which has the structure of a table (see Figa and 5B) to determine the appropriate relationship between the size D of the element, and the first width r1 of the strip, using the size D of the element data B(1) is divided into picture elements (generated by the block 101 split image) as the key, so as to define the first width r1 of the strip (first pass)defined in accordance with the size D of the element.

Block 104 calculation of the first array of coefficients of the filter calculates a first array of filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, using the first width r1 of the strip defined by the block 102 determining the first bandwidth.

Unit 105 to generate the data of the filtered image and exposes every data B(1) is divided into image elements generated by the block 101 split image) to the filtering process using the first array of filter coefficients, which is calculated by the block 104, the calculation of the first array of filter coefficients so as to generate data B(r1) of the first filtered divided into image elements.

Block 106 calculation of the distribution coefficient compares the data B(r1) of the first filtered divided into elements of the image data B(1) )is built on elements of the image and measures the value of P(r1), which is information about the errors and the ratio PSNR data B(r1) of the first filtered divided into image elements. Block 106 calculation of the distribution coefficient calculates the coefficient of X distribution based on the value of P(r1) by using the following formula:

X = G/P(r1)

where G is a constant that can be set to a 51.2", as shown in figure 1.

Block 107 table selection determine the optimal bandwidth selects one of the provided tables 108 determine the optimal bandwidth associated with the dimension D of the feature and the target ratio PSNR, and the table has a table structure shown in Figure 3, and corresponds to the dimension D of the data element B(1) is divided into picture elements (generated by the block 101 split image) and the target ratio PSNR, which is set by the user. Block 107 table selection determine the optimal bandwidth gives the ID assigned to the selected table.

With A_{i}and B_{i}specified in the table 108 to determine the optimal bandwidth, we have the following relations:

0<A_{1}<A_{2}<A_{3}<...<A_{n-2}<A_{n-1}

0<B_{1}<B_{2}<B_{3}<...<B_{n-2}<B_{n-1}<B_{n}<1

Block 109 determine the optimal Shi the ins bandwidth determines the optimal width of the r2 band (second pass) by reference to table 108 determine the optimal bandwidth, which is selected by the block 107 table selection determine the optimal bandwidth using ratio X occupancy (calculated by the block 106 calculation of the distribution coefficient) as the key.

Block 110 calculation of the optimal array of coefficients of the filter computes the array of optimal filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, using the optimum width r2 of the strip, which is defined by the block 109 determine the optimal bandwidth.

Unit 111 to generate the data of the filtered image and exposes every data B(1) is divided into picture elements (generated by the block 101 split image) to a filtering process using an array of optimal filter coefficients calculated by the calculation unit 110 of the array of optimal filter coefficients in order to generate data B(r2) optimum filtered divided into image elements and save them in block 112 of the accumulated data.

When all the data B(1) is divided into image elements, formed by block 101 split image, processed, all data B(r2) optimum filtered divided into image elements stored in block 112 data accumulation. the compliance with this unit 113 synthesis data of the filtered image and synthesizes the stored data and generates data B(r2)_all optimum filtered image data for the original image, entered through block 100 input original image data.

6 illustrates a block diagram of the sequence of operations performed by the device 1 forming the optimum filtered image of this variant implementation, performed as described above.

The processes performed by the device 1 forming the optimum filtered image, will be explained in detail in accordance with the flowchart of the sequence of operations.

As shown in the block diagram of the sequence of operations figure 6, when the device 1 forming the optimum filtered image receives the request data about the formation of the optimum filtered image relative to the image (which is the target of processing, and generates a video image), data B(1)_all the original image for which to derive the optimum filtered image is entered into the device (see the first step S100).

In the next step S101, based on the dimension D of the element or the number of E parts division, which are secured in advance the desired value, the input data B(1)_all the original image are separated, in order to generate data B(1) is divided into the elements of the image to the original image.

In the next step S102 to the table 103 definitions per the th bandwidth, which has the structure of a table (see Figa and 5B) to determine the appropriate relationship between the size D of the element, and the first width r1 of the strip, make use of the dimension D of the element data B(1) is divided into picture elements (1) as the key, so as to define the first width r1 of the strip (first pass)defined in accordance with the size D of the element.

If the size D of the element data B(1) is divided into image elements used in the device 1 for generating optimal filtered image of this variant implementation, limited to a predetermined size, the table 103 determining the first bandwidth is not required, and is determined by the first width r1 of the strip, which is set in advance in accordance with a fixed size.

In the next step S103, it calculates the first array of filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, using a specific first width r1 of the strip.

In the next step S104 selects some data B(1) is divided into image elements that have not yet been processed, and in the next step S105, the selected data B(1) is divided into picture elements are subjected to a filtering process using the calculated array of the first coefficient is icients filter, in order to generate data B(r1) of the first filtered divided into image elements.

In the next step S106, the selected data B(1) is divided into picture elements are compared with the generated data B(r1) of the first filtered divided into picture elements, and the measured value of P(r1), which is a data error and the ratio PSNR generated data B(r1) of the first filtered divided into image elements. Then calculate the coefficient of X distribution based on the value of P(r1) using the following formula:

X = G/P(r1) of Formula (1)

where G is a constant that can be set to a 51.2", as shown in figure 1.

In the next step S107 selects one of the provided tables 108 determine the optimal bandwidth associated with the dimension D of the feature and the target ratio PSNR, and the table has a table structure shown in Figure 3, and corresponds to the dimension D of the data element B(1) is divided into image elements and the target ratio PSNR, which is set by the user.

The above selection table 108 determine the optimal bandwidth can be made in advance.

In addition, if the size D of the element data B(1) of the original image used in the device 1 forming the optimal otherthrow the frame image of this variant implementation, limited to a predetermined size, it is not necessary to provide a table 108 determine the optimal bandwidth associated with the dimension D of the feature and the target ratio PSNR, and provides many tables 108 determine the optimal bandwidth associated with values for the target relations PSNR.

In addition, if the size D of the element data B(1) of the original image used in the device 1 for generating optimal filtered image, limited to a predetermined size, and the target ratio PSNR used in the device 1 for generating optimal filtered image, also limited to a predetermined value, it is not necessary to provide a table 108 determine the optimal bandwidth associated with the dimension D of the feature and the target ratio PSNR, and provides the only table 108 determine the optimal bandwidth.

In the next step S108 determines the optimal width r2 of the strip (second pass) by reference to the selected table 108 determine the optimal bandwidth using the calculated coefficient of X host as the key.

In the next step S109, it calculates the optimal array of filter coefficients for real is implementing frequency characteristics, corresponding to the limitation of bandwidth, using a specific optimal width r2 of the strip.

In the next step S110, the selected data B(1) is divided into picture elements again are filtered using the calculated optimal array of filter coefficients so as to generate data B(r2) optimum filtered divided into image elements and save them in block 112 of the accumulated data.

In the next step S111 is determined whether the selected portion of the data B(1) is divided into image elements. If it is determined that were chosen are not all part of the data B(1) is divided into image elements, the processing returns to step S104.

On the contrary, if in step S111 it is determined that you have selected all the parts of the data B(1) is divided into image elements, the processing goes to step S112 (. At step S112 (all part of the data B(r2) optimum filtered divided into image elements are synthesized to generate and output of data B(r2)_all optimum filtered image having the same size as the source image data. Then processing is completed.

Next will be specifically explained above-described processing.

Specifies that the original image size is 1920×1080; size D, item is 32×18; number E parts RA the division is 60 and the horizontal, and vertical directions; the value Ptgt relations PSNR (that is, the target ratio PSNR) is equal to 36 dB; and the value of G in the formula (1) is 51,2.

First will be explained the process of determining the first bandwidth.

Size D, item is entered in block 102 determining the first bandwidth, and is determined by the first width r1 strips (for example, 0,7) for dimension D of the element using table 103 determining the first bandwidth, which is provided in advance to the block 102 determining the first bandwidth.

Then derive B(0,7) of the first filtered divided into image elements using a first array of filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth with r1=0,7, and the measured value of P(0,7), which is the ratio of the PSNR for the data B(0,7) of the first filtered divided into image elements. Then calculate the coefficient of X distribution using the formula (1).

Next will be explained the process of determining the optimal bandwidth.

Values, for example, D=32×18 and Ptgt=36, entered in block 107 table selection determine the optimal bandwidth, and selects one of the tables 108 determine the optimal bandwidth, which is provided in advance what Locke 107 table selection determine the optimal bandwidth, moreover, the selected table matches the entered values and has the table structure shown in Figure 3.

Then we can determine the optimal width r2 of the strip corresponding to the previously calculated rate X occupancy, by reference to the selected table 108 determine the optimal bandwidth.

For example, if P(0,7)=45, X=1,14 in accordance with formula (1). Therefore, if A_{n-2}≤1,14<A_{n-1}then the optimum width r2 of the strip is defined as B_{n-1}. While A_{i}and B_{i}respectively satisfy the following conditions.

0<A_{1}<A_{2}<A_{3}<...<A_{n-2}<A_{n-1}

0<B_{1}<B_{2}<B_{3}<...<B_{n-2}<B_{n-1}<B_{n}<1

Data B(r2) optimal divided into image elements are formed using an array of optimal filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth using the optimum width r2 of the strip.

Such processes are repeated this number of times, which corresponds to the number of parts of the separation, that is applied to (60×60=) 3600 elements. 3600 pieces of data B(r2) optimal divided into image elements finally synthesized, so that the data of the optimum filtered image were obtained as eventually the x output.

In addition, many tables 108 determining the optimum bandwidth corresponding to the different values Ptgt, can be prepared in advance, in order to perform a filtering process to implement arbitrary image quality using the present invention.

As described above, in accordance with the first variant embodiment it is possible to obtain elements having almost equal relationship PSNR for any image, and thus to form a filtered image whose image quality is homogeneous in each frame and all areas in the frame are almost equal image quality.

(2) the Second variant implementation

7 illustrates an example of the structure of the device 11 forming the optimum filtered image as a second variant implementation of the present invention.

Compared to device 1 forming the optimum filtered image (see Figure 4) of the first variant of implementation, the device 11 forming the optimum filtered image of the second variant implementation has no unit 101 split image, but additionally includes block 120 job of dividing elements. In addition, the device 11 forming the optimum filtered image block 102α determine what erway bandwidth, block 105α generate the data of the filtered image and the block 106α calculate the distribution coefficient respectively perform processes other than those that perform block 102 determining the first bandwidth, the unit 105 to generate the data of the filtered image and the block 106 calculate the distribution device 1 forming the optimum filtered image of the first variant implementation.

Based on the size D of an element or a number of parts of E division, which are set in advance, the block 120 job of dividing elements specifies the elements of the virtual separation of data B(1)_all the original image entered through block 100 input original image data. Image data of each element of virtual separation correspond to each data B(1) is divided into the elements of the image to the original image, which have been explained in the first embodiment.

Block 102α determining the first bandwidth refers to table 103 determining the first bandwidth, which has the structure of a table (see Figa and 5B) to set the corresponding relationship between the size D of the element, and the first width r1 of the strip, using the size D of the element for the virtual elements of the division (specified by the block 120 jobs e the elements of separation) as a key, in order to define the first width r1 of the strip (first pass)defined in accordance with the size D of the element.

Block 105α generate the data of the filtered image and puts the data B(1)_all the original image (introduced by block 100 of the input original image data) to the filtering process using the first array of filter coefficients, which is calculated by block 104 calculate the first filter coefficients in order to generate data B(1)_all(r1) filtered image.

For each element of virtual separation specified by block 120 job of dividing elements block 106α calculate the distribution coefficient, comparing the portion of the image data, which is data B(1)_all(r) of the first filtered image and placed in the corresponding element of the split, with part of the image data, which is data B(1)_all the original image and placed in the corresponding element of separation, and measures the value of P(r1), which is a data error and the ratio PSNR part of the image data, which is data B(1)_all the first filtered image and placed on the corresponding element of separation. Block 106α calculate the distribution coefficient calculates the coefficient of X distribution based on the value of P(r1) IP is the use of the following formula:

X=G/P(r1)

Fig illustrates a block diagram of the sequence of operations performed by the device 11 forming the optimum filtered image of this variant implementation, performed as described above.

In accordance with the flowchart of the operational sequence of the processes performed by the device 11 forming the optimum filtered image, will be explained in detail.

As shown in the block diagram of the sequence of operations on Fig, when the device 11 forming the optimum filtered image receives the request data about the formation of the optimum filtered image relative to the image (which is the target of processing, and generates a video image), data B(1)_all the original image for which to derive the optimum filtered image is entered into the device (see the first step S200).

In the next step S201 to the table 103 determining the first bandwidth, which has the structure of a table (see Figa and 5B) to set the corresponding relationship between the size D of the element, and the first width r1 of the strip, make use of the dimension D of the element of the elements of the virtual separation as a key so as to define the first width r1 of the strip (first pass)defined in accordance with the size of the om-D element.

In the next step S202, it calculates the first array of filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, using a specific first width r1 of the strip.

In the next step S203, the input data B(1)_all the original image subjected to the filtering process using the calculated first array of filter coefficients so as to generate data B(1)_all(r1) of the first filtered image.

In the next step S204 one element separation, which has not yet been processed is selected from among the virtual set of elements of the division.

In the next step S205, the portion of the image data, which is data B(1)_all(r1) of the first filtered image and placed on the selected element separation, compared with a part of the image data, which is data B(1)_all the original image and placed on the selected element separation, and the measured value of P(r1), which is the ratio of the PSNR of the above-mentioned part of the image data for the data B(1)_all(r1) of the first filtered image. Then the coefficient of X distribution is calculated based on the value of P(r1) with the following formula:

X = G/P(r1) of Formula (1)

where G is a constant that can be set to a 51.2", as shown in figure 1.

In the next step 206 selects one of the tables 108 determine the optimal bandwidth, which is provided in connection with the dimension D of the feature and the target ratio PSNR, and the table has a table structure shown in Figure 3, and corresponds to the dimension D of the element for the specified virtual elements of separation and the target ratio PSNR, which is set by the user.

In the next step S207 determines the optimal width r2 of the strip (second pass) by reference to the selected table 108 determine the optimal bandwidth using the calculated coefficient of X host as the key.

In the next step S208, it calculates the optimal array of filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, using a specific optimal width r2 of the strip.

In the next step S209, the portion of the image data, which is data B(1)_all the original image and placed on the selected element separation, again subjected to filtration using the calculated optimal array of filter coefficients so as to generate data B(r2) optimal divided into image elements and save them in block 112 of the accumulated data.

In the next step S210 determines whether the selected all elements of the division. If it is determined that not all elements of the division were selected, the processing returns to step S204.

Conversely, if at step S210 it is determined that you have selected all the elements of separation, processing proceeds to step S211. At step S211 all parts of the data B(r2) optimal divided into image elements are synthesized to generate and output of data B(r2)_all optimum filtered image having the same size as the source image data. Then processing is completed.

In the structure shown in Fig.7, the data B(1)_all(r1) of the first filtered image, generated by the block 105α generate the data of the filtered image, the virtual divided into elements. However, this separation can be accomplished in fact.

As described above, similarly to the first variant embodiment of the second variant implementation is also possible to obtain elements having almost equal relationship PSNR for any image, and, thus, to form a filtered image whose image quality is homogeneous in each frame and all areas in the frame are almost equal image quality.

(3) a Third option exercise

Fig.9 illustrates an example of the structure of the device 12 forming the optimum filtered image as the third alternative implementation of the present invention.

Compared with the device 1 for generating optimal otfit is consistent image (see 4) the first variant implementation of the device 12 forming the optimum filtered image of the third variant implementation additionally includes block 130 comparison of the optimal bandwidth and the block 131 of the adjustment to the optimal bandwidth.

Block 130 comparison of the optimal bandwidth compares the optimal width of the r2 band (defined by the block 109 determine the optimal bandwidth) data B(1) is divided into image elements for the target element processing with optimal bandwidth data B(1) is divided into image elements for the peripheral element of the target element; computes the difference between the compared values; and determines whether the difference more than a predetermined threshold Sth1 or equal to.

If the block 130 comparison of the optimal bandwidth determines that the difference is greater than or equal to the threshold Sth1, block 131 adjusting the optimal bandwidth adjusts the optimal width r2 of the strip defined using block 109 determine the optimal bandwidth, the value of r3, and thereby decreases the difference. On the contrary, if it is determined that the difference is less than the threshold Sth1, block 131 adjusting the optimal bandwidth determines Thu the optimal width r2 of the strip, determined using block 109 determine the optimal bandwidth is used without modification.

Figure 10 and 11 show a flowchart of the sequence of operations performed by the device 12 forming the optimum filtered image of this variant implementation, performed as described above.

The processes performed by the device 12 forming the optimum filtered image, will be explained in detail in accordance with the flowchart of the sequence of operations.

When prompted about the formation of a data optimum filtered image relative to the image (which is the target of processing, and generates a video image), device 12 forming the optimum filtered image and performs the same processes (steps S300 through S308), as in steps S100 through S108 of the flowchart of the sequence of operations shown in Fig.6, in order to determine the optimal width of the r2 band (second pass) for the selected data B(1) is divided into image elements.

In the next step S309, the optimum width r2 of the band (as determined in step S308), the target of the processing elements is compared with the already computed the optimal bandwidth element placed around the target element; calculates a difference between sravan the mi values; and determines whether the difference more than a predetermined threshold Sth1 or equal to.

In accordance with the above determination, if it is determined that the difference between the optimal width r2 strips the target of the processing elements and the optimal bandwidth compared peripheral element is greater than or equal to the threshold Sth1, the processing proceeds to step S310, where the optimal width r2 of the strip, as determined in step S308, is adjusted to a value r3, whereby the difference decreases.

On the contrary, if it is determined that the difference between the optimal width r2 strips the target of the processing elements and the optimal bandwidth of the peripheral element is less than the threshold Sth1, step S310 is performed, and the optimal width r2 of the strip, as determined in step S308, is used without changes.

The following steps S311 through S314 is executed the same processes as in the steps S109 through S112 (a flowchart of the sequence of operations shown in Fig.6, in order, were formed and derived data B(r2)_all optimum filtered image having the same size as the source image data.

Fig-15 show examples of the adjustment process for optimal bandwidth, performed by the block 131 of the adjustment to the optimal bandwidth.

If the Central nervous system, the th element among the nine elements, which with the help of block 109 determine the optimal bandwidth was appointed optimal bandwidth is in order to correct, block 131 adjusting the optimal bandwidth adjusts the optimal bandwidth (defined using block 109 determine the optimal bandwidth, as shown in Fig-14.

Thus, as shown in Fig, the optimal bandwidth can be adjusted to (i) the value equal to the value assigned to the upper and lower elements relative to the target processing element, (ii) the average of these values, or (iii) the value obtained additional gain values ±β (0<β<1) adjusted to the value shown in the above item (i) or (ii).

In addition, as shown in Fig, the optimal bandwidth can be adjusted to (i) the value equal to the value assigned to the left and right elements relative to the target processing element, (ii) the average of these values, or (iii) the value obtained additional gain values ±β (0<β<1) adjusted to the value shown in the above item (i) or (ii).

In addition, as shown in Fig, the optimal bandwidth can be correcti avana to (i) the value equal to the values assigned to the items placed diagonally relative to the target processing element, (ii) the average of these values, or (iii) the value obtained additional gain values ±β (0<β<1) adjusted to the value shown in the above item (i) or (ii).

In addition, as shown in Fig, the optimal bandwidth can be adjusted to the average value of eight peripheral elements of the target processing element or to the value obtained by adding the values ±β (0<β<1) to the average value.

Any of the mentioned above methods has a similar effect.

In accordance with such adjustment of boundary lines obtained from the filtering process applied to the elements is reduced, and thus the framework can be invisible.

If the above adjustment is individually applied to the top field and the bottom field in the television processing striped, similar effects can be obtained when the peripheral element, compared to the target processing element belongs to either the same field as the first target element, or field, other than the target element.

This adjustment can also be accomplished by extending the range of the distribution coefficient in the table is CE 108 determine the optimal bandwidth, having the table structure shown in Figure 3.

For example, in figure 3, if the optimal bandwidth of 0.8 is assigned to the factor of 1.5≤X<1,6, and the optimal bandwidth of 0.9 is assigned to the coefficient of 1.6≤X<1,7, these conditions can be changed on condition that the optimal bandwidth 0,85 assigned to a factor of 1.5≤X<1,7.

In accordance with a third variant embodiment of the deterioration of subjective image quality, such as block distortion can be reduced, and thus largely preserved the original objective quality of the image.

(4) the Fourth option exercise

Fig illustrates an example of the structure of the device 13 forming the optimum filtered image as a fourth variant of implementation of the present invention.

Compared with the device 12 forming the optimum filtered image (see Fig.9) the third alternative implementation of the device 13 forming the optimum filtered image of the fourth version of the implementation additionally includes block 140 identifying the elements with movement and block 141 additional adjustment of the optimal bandwidth.

Block 140 identifying the elements with movement determines whether the movement of the element, whichhas data B(1) is divided into picture elements (as processing purposes) original image (that is, the movement is detected within the element). If it is determined that the corresponding element has no motion, the block 140 identifying the elements with the movement passes the result processing unit 131 adjusting the optimal bandwidth directly to the block 110 calculation of the optimal array of coefficients of the filter.

If using block 140 definitions of the elements of movement is determined that the corresponding element is movement, block 141 additional adjustment of the optimal bandwidth additionally adjusts the optimal bandwidth adjusted unit 131 adjusting the optimal bandwidth (which may not re-adjust the optimal bandwidth).

The block 105 forming the filtered image data, the block 106 calculation of the distribution coefficient, block 109 determine the optimal bandwidth, block 130 comparison of the optimal bandwidth and block 131 adjusting the optimal bandwidth form part of intraframe image processing; and a calculation unit 110 of the array of optimal filter coefficients, the block 140 identifying the elements with movement and block 141 additional adjustment of the optimal bandwidth form part of interframe image processing.

Fig and 18 show a block diagram of the sequence of operations, executable device 13 forming the optimum filtered image of this variant implementation, performed as described above.

The processes performed by the device 13 forming the optimum filtered image, will be explained in detail in accordance with the flowchart of the sequence of operations.

As shown in the block diagram of the sequence of operations on Fig, when receiving request data about the formation of the optimum filtered image relative to the image (which is the target of processing, and generates a video device 13 forming the optimum filtered image and performs the same processes (steps S400 through S408), as in steps S300 through S308 flowchart of the sequence of operations shown in Figure 10, in order to determine the optimal width of the r2 band (second pass) for the selected data B(1) is divided into image elements.

In the next step S409, a certain optimum width r2 of the strip is adjusted based on the intraframe image processing. This adjustment process certain optimal width r2 of the strip based on the intraframe image processing is performed similarly to the process at steps S309 and S310 flowchart of the sequence of operations shown in Figure 10.

Thus, primulina width r2 of the band (as determined in step S408), assigned data B(1) is divided into image elements for the target processing element, is compared with the optimal bandwidth assigned to the data B(1) is divided into image elements for an item placed around the target element; calculates a difference between the compared values; and determines whether the difference more than a predetermined threshold Sth1 or equal. If it is determined that the difference is greater than or equal to the threshold Sth1, the optimum width r2 of the strip, as determined in step S408, is adjusted to a value r3, whereby the difference decreases. On the contrary, if it is determined that the difference is less than the threshold Sth1, it is also determined that the optimal width r2 of the strip, as determined in step S408, is used without changes.

In the next step S410, the optimum width of r3 (or r2) band adjusted at step S409, additionally adjusted based on interframe image processing, which will be explained using a flowchart of the sequence of operations shown in Fig.

The following steps S411 through S414 are the same processes as in steps S311 through S314 flowchart of the sequence of operations shown in Figure 10 and 11, so that were formed and derived data B(r2)_all optimum filtered image having the same size as Yes the data of the original image.

Next, with reference to the block diagram of the sequence of operations shown in Fig, will be explained the process of adjustment to the optimal bandwidth based on interframe image processing performed at the step S410.

After the optimal width r2 of the strip, as determined in step S408, adjusted at step S409-based intraframe image processing (see the block diagram of the sequence of operations on Fig), the processing proceeds to the block diagram of the sequence of operations on Fig. At the first stage S500 is calculated for the full amount of the pixel value of the target processing element, to which was applied intra-frame image processing. At next step S501 is calculated for the full amount of pixel values of an element that belongs to the previous frame (i.e. precedes in time the frame of the target element) and spatial identical to a target processing element.

In the above-mentioned process can be calculated for the full amount of the pixel value data B(1) is divided into picture elements (that is, original image data) or can be calculated for the full amount of the pixel value data B(r1) of the first filtered image (that is, data of the filtered image).

In the next step S502, it calculates the difference between the full amount computed on this is e S500, and full the amount computed in step S501. In the next step S503 to determine whether the difference more than a predetermined threshold Sth2 or equal to.

In accordance with the above determination, if it is determined that the difference between the full amount of the pixel value of the target element processing and full sum of pixel values of an element that belongs to the previous frame and the spatial identical to a target processing element, is greater than or equal to the threshold Sth2, it is determined that the target element processing has movement, and the processing proceeds to step S504. At step S504, the optimum width of r3 (or r2) band adjusted at step S409, optionally adjusted to a value of r4, so as to reduce the optimal width of r3 (or r2) bands, and the processing proceeds to step S411 flowchart of the sequence of operations shown in Fig.

For example, the optimal width of r3 (or r2) band adjusted at step S409, is multiplied by the weight W, which is less than unity (i.e. 0<W<1), so as to further adjust the value of r3 (or r2) is r4, and then processing proceeds to step S411 flowchart of the sequence of operations shown in Fig.

In accordance with the above-described adjustment may be significantly limited by the high frequency component in the element with the movement.

Naproti is, if the above definition is determined that the difference between the full amount of the pixel value of the target element processing and full sum of pixel values of an element that belongs to the previous frame and the spatial identical to a target processing element, is less than the threshold Sth2, the process of step S504 is performed and determined that the optimal width of r3 (or r2) band adjusted at step S409, is used without change. Then, the processing proceeds to step S411 flowchart of the sequence of operations shown in Fig.

Although the block diagram of the sequence of operations shown in Fig calculates the total values of the pixels in the element can be computed average value of the pixels in the element.

In contrast to the first, second and third variants of the embodiment in accordance with the fourth alternative embodiment may be largely preserved the original subjective image quality, although the objective quality of the image varies, so it is possible to prevent the deterioration of subjective image quality and improve the corresponding coding efficiency.

(5) the Fifth option exercise

Fig illustrates an example of the structure of the device 14 forming the optimum filtered image as the fifth option implemented the I of the present invention.

Compared with the device 12 forming the optimum filtered image (see Fig.9) the third alternative implementation of the device 14 forming the optimum filtered image fifth variant of the implementation additionally includes block 150 to determine the high-frequency elements, block 151 measurement of high-frequency elements with movement and block 152 additional adjustment of the optimal bandwidth.

Block 150 to determine the high-frequency elements determines whether the element that has the data B(1) is divided into the elements of the original image (as processing purposes), element, characterized by high-frequency component. If it is determined that the corresponding element is not such an element with a high-frequency component, the block 150 to determine the high-frequency elements transmits the result of processing block 131 adjusting the optimal bandwidth directly to the block 110 calculation of the optimal array of coefficients of the filter.

If the block 150 to determine the high-frequency elements is determined that the corresponding element is a high-frequency component unit 151 measurement of high-frequency elements with movement measures the number of cells with a high-frequency component in the frame, which is anaglesic target processing element, as well as the number of cells with a high-frequency component in the previous frame (that is prior in time to the frame of the target element). Based on the results of the measurement unit 151 measurement of high-frequency elements with movement determines whether the target processing motion (i.e. motion is detected within the element). If it is determined that the corresponding element has no motion, the block 151 measurement of high-frequency elements with the movement passes the result processing unit 131 adjusting the optimal bandwidth directly to the block 110 calculation of the optimal array of coefficients of the filter.

If the block 151 measurement of high-frequency elements of the movement is determined that the corresponding element has a movement that is finally determined that the target processing element is an element with a high-frequency component with the movement of the block 152 additional adjustment of the optimal bandwidth additionally adjusts the optimal bandwidth adjusted unit 131 adjusting the optimal bandwidth (which may not re-adjust the optimal bandwidth).

The block 105 forming the filtered image data, the block 106 calculate the ratio of the distribution, block 109 determine the optimal bandwidth, block 130 comparison of the optimal bandwidth and block 131 adjusting the optimal bandwidth form part of intraframe image processing; and a calculation unit 110 of the array of optimal filter coefficients, the block defining the high-frequency elements, block 151 measurement of high-frequency elements with movement and block 152 additional adjustment of the optimal bandwidth form part of interframe image processing.

Similarly, the device 13 forming the optimum filtered image of the fourth version of the exercise device 14 forming the optimum filtered image fifth variant of the implement having the above described structure, executes a block diagram of the sequence of operations shown in Fig. However, unlike device 13 forming the optimum filtered image, the device 14 forming the optimum filtered image and executes the adjustment process for optimal bandwidth based on interframe image processing (see step S410), in accordance with the flowchart of the sequence of operations shown in Fig.

Next, with reference to the block diagram of the sequence of operations, pokazanno is on Fig, explained the process of adjustment to the optimal bandwidth based on interframe image processing performed by the device 14 forming the optimum filtered image of this variant implementation.

In the device 14 forming the optimum filtered image of how the optimal width r2 of the strip, as determined in step S408, adjusted at step S409-based intraframe image processing (see the block diagram of the sequence of operations on Fig), the processing proceeds to the block diagram of the sequence of operations shown in Fig. At the first stage S600 retrieves the coefficient of X(n,m) placement of the target processing element, to which was applied intra-frame image processing.

The coefficient of X(n,m) was calculated at step S406 flowchart of the sequence of operations shown in Fig, where n is the number of the frame that owns the target processing element, and m is the number of target processing element.

In the next step S601 is determined whether more extracted coefficient of X(n,m) placement than a predefined threshold Xth. If it is determined that the extracted coefficient of X(n,m) allocation is less than or equal to the threshold Xth, it is determined that the target processing element is not an element, characterized by high-frequency sostavlyajushie is, and that the optimum width of r3 (or r2) band adjusted at step S409 (a flowchart of the sequence of operations on Fig), is used without change and without performing the following processes. In line with this, processing moves to step S411 flowchart of the sequence of operations shown in Fig.

On the contrary, if at the step S601 is determined that the extracted coefficient of X(n,m) placement more threshold Xth, it is determined that the target processing element is an element with a high-frequency component, and the processing proceeds to step S602. At step S602, based on the ratio X of every element in the frame belongs to the target processing element, calculates the number M(n) elements of a high-frequency component belonging to the corresponding frame.

In the next step S603 on the basis of the coefficient X of placement for each element in a previous frame immediately preceding the frame belongs to the target processing element, calculates the number M(n-1) elements of a high-frequency component belonging to the previous frame.

In the next step S604, it calculates the difference |M(n)-M(n-1)| between the number M(n) elements computed in step S602, and the number M(n-1) elements computed in step S603, and determines whether the difference more predefined Mth. If it is determined that the difference is less than or equal to the threshold Mth, it is also determined that the frame belongs to the target processing element, does not show the movement and the target processing element is not an element with the movement. Therefore, it is also determined that the optimal width of r3 (or r2) band adjusted at step S409 (a flowchart of the sequence of operations on Fig), is used without change and without performing the following processes. In line with this, processing moves to step S411 flowchart of the sequence of operations shown in Fig.

On the contrary, if at the step S604 is determined that the difference between |M(n)-M(n-1)| between the number M(n) elements computed in step S602, and the number M(n-1) elements computed in step S603, more threshold Mth, it is also determined that the frame belongs to the target processing element, manifests the presence of motion, and the target processing element is the element with the movement. Therefore, the processing goes to step S605, where the optimal width of r3 (or r2) band adjusted at step S409, optionally adjusted to a value of r4, so as to reduce the value of r3 (or r2). Then, the processing proceeds to step S411 flowchart of the sequence of operations shown in Fig.

For example, the optimal width of r3 (or r2) band adjusted at step S409, amnuaysin weight W, which is less than unity (i.e. 0<W<1), so as to further adjust the value of r3 (or r2) is r4, and then processing proceeds to step S411 flowchart of the sequence of operations shown in Fig.

In accordance with the above-described adjustment may be significantly limited by the high frequency component in the element with a high-frequency component with the movement.

Next will be specifically explained above processing.

It is assumed that the optimal width of r2 (or r3) of the band, which was obtained using the process unit 131 adjusting the optimal bandwidth, equal to 0.9; the number n of the frame that owns the target processing element equal to 5; the number m of the corresponding element is equal to 1000; the calculated distribution coefficient X(5,1000) more 1,9; the threshold Xth (=1,9) secured to the block 150 to determine the high-frequency elements; threshold Mth (=15) is secured to the block 151 measurement of high-frequency elements with the motion; and the weight coefficient W (=0,7) secured to the block 152 additional adjustment the optimal bandwidth.

First, the block 150 to determine the high-frequency element determines that the target processing element is an element with a high-frequency component, since the distribution coefficient X(5,1000)computed by block 106 calculate the CoE is ficient, more threshold Xth (=1,9).

Then the block 151 measurement of high-frequency elements with movement, calculates the number of elements that satisfy the condition X(5)>Xth(=1,9), among the coefficients of X(5) placement computed by block 106 calculation of the coefficients of occupancy, and calculates a fourth frame immediately preceding the fifth frame, the number of elements that satisfy the condition X(4)>Xth(=1,9), among the coefficients of X(4) placement. Here it is assumed that the calculated values of M(5) and M(4) are respectively 11 and 35.

Then the block 151 measurement of high-frequency elements with movement calculates the difference |M(5)-M(4)| between the computed values of M(5) and M(4), that is |11-35|=24. In this case, 24 more than Mth (=15), and therefore finally determined that the target processing element is an element with a high-frequency component with the movement.

In line with this, the block 152 additional tuning bandwidth applies weighting to the optimal width r2 strips the target of the processing, that is, r4=r2×W (=0,7), and the optimal width of the r4 band equal to 0.63. It then executes the filtering process using the updated optimal bandwidth, in order to obtain the final data of the optimum filtered image.

As is clear from the above-described processing, when the each of the elements, corresponding to the condition M(5)=11, becomes the target of processing, the optimum width r2 of the strip is subjected to a similar weighting (i.e. r4=r2×W (=0,7)), and performs a filtering process using the updated optimal bandwidth, in order to obtain the final data of the optimum filtered image.

In accordance with the above-described processing ratio PSNR of each high-frequency element having a movement, becomes equal to 30 dB, and the PSNR of each high-frequency component having no motion, is equal to 40 dB, and the difference visually imperceptible.

In the case when M(5)=11 and M(4)=21, the condition |M(5)-M(4)|>Mth is not satisfied, and therefore is determined that the target element processing is a high-frequency element without movement. Therefore, the adjustment unit 152 additional adjustment of the optimal bandwidth is not needed, and the optimal width r2 of the strip becomes the optimal width of r4 bands unchanged. Then executes the corresponding filtering process in order to obtain data of the optimum filtered image.

The above value of M(n) varies depending on the threshold Xth, and the criteria for determining the presence or absence of movement is set depending on the threshold Mth. Therefore, the OIG Xth and Mth should be established with regard to the element size, etc.

Regarding the setting of the threshold Mth, instead of providing a fixed numerical value, can be expected Mth=E×0,1, i.e. in relation to the number of E parts division.

If you use this attitude Mth=E×0,1, compared with the previous frame of 10% of the number of elements with the high-frequency component changed from the elements with a high-frequency component on the elements with the low-frequency component, while the corresponding number of cells with low-frequency component is changed from the elements with the low-frequency component on the elements of a high-frequency component. Also in this case can be obtained effects similar to those obtained by providing a fixed numerical values.

In addition, although when comparing the number of elements with a high-frequency component of a previous frame immediately preceding the current frame, similar effects can be obtained by using another frame before or after the current frame.

In accordance with a fifth alternative embodiment of the high frequency component can be significantly limited in the area with movement. Therefore, unlike the first, second and third variants of the embodiment, and similarly to the fourth variant embodiment of the initial subjective quality is in the image can be greatly saved, although changes of the objective quality of the image, and it is therefore possible to prevent deterioration of subjective image quality and improve the corresponding coding efficiency.

Thus, in accordance with a fifth alternative embodiment may be significantly limited by the high frequency component with the movement. Therefore, the coding efficiency can be improved with less deterioration in subjective image quality compared with the fourth alternative embodiment.

(6) Of the present invention

As explained in the first or second embodiment, the present invention does not use a method in which the data B(r2)_all optimum filtered image are formed using a common optimal width r2 of the band for the whole frame corresponding to the video image, and generates data B(r2) optimum filtered image using the optimum width r2 of the band assigned to each item specified by separating the image data for the frame, and further generates data B(r2)_all optimum filtered image by synthesizing each data B(r2) optimum filtered image.

Therefore, only if the corresponding frame (see frame N on Fig) there is a region characterized by low frequency with the excitation, data optimum filtered image formed with the use of force filter, corresponding to the low-frequency component.

On the contrary, if the corresponding frame (see frame N+1 on Fig) are present and the area is characterized by low-frequency component, and an area characterized by high-frequency component data of the optimum filtered image are formed by separate use of force filter, corresponding to the low-frequency component, to the area with the low-frequency component and the strength of the filter corresponding to the high-frequency component, to the field of high-frequency component.

Therefore, in accordance with the present invention (i) each frame can have a uniform image quality and, thus, a uniform display, and (ii) the quality of the image within each frame can be homogeneous, and thus, the display within the frame can also be homogeneous.

Also in accordance with the present invention having the above-described effects, noises on the borders of the regions, caused by the filtering process applied to each area can be reduced, as described in the third embodiment, thereby reducing the deterioration of the subjective quality of the image.

Also in accordance with this the image is the group of with the above-described effects, the high-frequency component in the area having a motion, may be significantly restricted, as described in the fourth embodiment, or the high-frequency component in the field of high-frequency component having a motion may be significantly limited, as described in the fifth embodiment, and thereby improving the coding efficiency without causing deterioration of the subjective quality of the image.

Fig illustrates the results of an experiment performed to test the effectiveness of the present invention.

For comparison with the present invention in the above-mentioned experiment, the comparative data of the optimum filtered image were formed using a common optimal bandwidth for the whole frame. In addition, there were data in the optimum filtered image in accordance with a third variant embodiment, the subjective image quality which was substantially equal to the subjective image quality for comparative data the optimum filtered image. In addition, there were data in the optimum filtered image in accordance with the fourth alternative embodiment, the subjective image quality which was in C is uchitelnoj extent is subjective image quality for comparative data the optimum filtered image. The above-mentioned three pieces of data of the optimum filtered image were subjected to coding under the same conditions, and was calculated amount of code. Were then calculated reduction factor by comparing each number code number code obtained by encoding the source image data.

On Fig horizontal axis denotes the quantization parameter (QP)used for encoding, and the vertical axis denotes the reduction ratio of the amount of code.

On Fig (i) experimental data of a single frame have been received by the data generating optimum filtered image using the overall optimal bandwidth for the whole frame, (ii) experimental data of a single element based on movement were obtained by the data generating optimum filtered image in accordance with a third variant of the embodiment, and (iii) experimental data of a single element without movement" were obtained by the data generating optimum filtered image in accordance with the fourth alternative embodiment.

As for the experimental data, it can be confirmed that the amount of generated code can be significantly reduced with the use of the Finance of the present invention and thus can be largely preserved the original subjective image quality. Therefore, the effectiveness of the present invention could be confirmed.

Although the present invention has been explained in accordance with a variant embodiment with reference to the drawings, the present invention is not limited to variants of the incarnation.

For example, although it is assumed that the ratio PSNR is used as an example of the error information explained in the above embodiments, embodiments, similar effects can be obtained by using the standard error, variance, etc. that includes error information for the respective pixels.

In addition, explained above options assume embodiment is an example in which the original image size is 1920×1080 and the element size is 32×18. However, the table 103 determining the first bandwidth, which sets the first width r1 of the bands corresponding to the different sizes of the original image data and the size of the element can be prepared in advance and provided to block 102 to determine the bandwidth in order to apply the present invention to images having any desired size.

In addition, explained above options embodiments assume that each element has a rectangular shape. However, the shape of each element is also not limited, and a similar effect is you can be obtained by using the forms, different from the rectangle (for example, cross, triangle or circle).

Explained above options embodiments also suggest that the number of parts of the split image is the same in the horizontal and vertical directions. However, similar effects can be obtained even when different parts of the separation (for example, E_{1}and E_{5}respectively assigned to the horizontal and vertical directions.

In addition, the fourth and fifth options embodiments assume that the weighting coefficient has the same value in the horizontal and vertical directions. However, similar effects can be obtained even when using different values.

Explained above options embodiments also suggest that the first bandwidth and the optimal bandwidth, each are the same in the horizontal and vertical directions. However, similar effects can be obtained even when different values (for example, B_{1}and B_{5}(to limit the bandwidth)), respectively, are assigned to the horizontal and vertical directions, in order to positively use the following phenomenon: in the video, showing a distant view of the landscape or a large truck, there is a pain the e changes the brightness in the vertical direction compared to the horizontal direction,
because there is gravity in the vertical direction.

Although the above-described variants of the embodiment is not provided guidance on what filter you use, you may use a digital filter with 7 taps, and similar effects can be obtained using a different number of taps.

In addition, not subject to any special restriction on a method of designing a digital filter for implementing the specified limit of the band. For example, the desired appearance of the frequency characteristics may be subjected to inverse Z-transform in order to obtain and to design an array of filter coefficients of the digital filter having the appropriate frequency response.

Also in the above-described embodiments, the embodiment is 51,2 is used as the value of G in the formula used to calculate the coefficient of X host. However, the value of G depends on the characteristics of the digital filter must be modified accordingly, when using a different digital filter.

Although the above-described variants of the embodiment are not provided special consideration, treatment strips may be applied not only to the component of the brightness, but also to the component of the color difference. In this case, the coding efficiency can be complete is Ino improved.

In addition, in embodiments of the incarnation is only the threshold for the lower limit, for example, X(n,m)>Xth. However, similar effects can be obtained by setting a threshold for the upper limit.

In addition, each explained above fourth and fifth variants of the embodiment performs intraframe processing image and interframe image processing. However, similar effects can be obtained when performing any processing of intra-frame image processing and frame-to-frame image processing.

Although there was provided an explanation of the combination between the above-described variant embodiments, any combination between the variants of embodiment is possible, and similar effects can be obtained, even when you change the order of execution of the relevant processes.

__INDUSTRIAL APPLICABILITY__

In accordance with the present invention the adaptive filtering process for images that form a video image, can be implemented without coding and without repeating the processing with regard to the frequency distribution in the frame or between frames of images, and thereby effectively formed a filtered image that has the specified value assessment of the quality of the image.

1. The rendering of containing phases in which:br/>
divide the target image processing, which forms a video image into many regions of separation;

determine the first bandwidth applied to the areas of separation;

calculate a first array of filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, by using the first bandwidth;

form data is filtered divided image for each area of the division, by exposure of the image data of each region separating the filtering process using the first array of filter coefficients;

output for each area of the division is information about the error between the image data of each area separation data and the filtered divided image and calculate the distribution coefficient used for determining an optimum bandwidth, based on the extracted values;

determine for each area of the separating optimum bandwidth corresponding to the distribution coefficient;

calculate for each region separating the array of optimal filter coefficients for implementing frequency characteristics,

corresponding to the limitation of bandwidth, using a specific optimal bandwidth etc the transmission;

form data optimum filtered divided image of each region dividing by the exposure of the image data of each region separating the filtering process using an array of optimal filter coefficients; and

synthesize data optimum filtered divided image of each area of separation.

2. The video processing method according to claim 1, in which: the step of determining the first bandwidth, the first bandwidth is determined based on the size of each field division.

3. The video processing method according to claim 1, additionally containing phases in which: compare for each area of the separation of certain optimal bandwidth with optimal bandwidth of the peripheral area of separation around this area separation; and adjust a certain optimal bandwidth based on the comparison result.

4. The video processing method according to claim 1, further comprising stages, which determine whether the motion image data of each region, division, using the image data of the frame before or after the frame area or image data frames before and after the frame region; and correcting certain optimal bandwidth of each region is STI division, for which it is determined that the image data area splitting have movement.

5. The video processing method according to claim 4, in which the step of determining whether the moving image data of each area separation is performed by evaluating the variations of the pixel values from each frame to the given frame.

6. The video processing method according to claim 4, in which the step of determining whether the moving image data of each area separation is performed by determining whether the movement data of the image region separation, and is characterized by high-frequency component; and a phase adjusting certain optimal bandwidth is performed by adjusting a certain optimal bandwidth each area of separation, which determined that these images have motion, and characterized by high-frequency component.

7. The video processing method according to claim 6, in which: the step of determining whether the moving image data of each region, division, and characterized by high-frequency component is performed by stages, in which: determine whether the value is an attribute of the image data area splitting that image data are characterized by high-frequency component; the assessment of the variation in the number of areas of separation, which shows that the image data are characterized by high-frequency component, from each frame to the given frame.

8. The video processing method according to claim 1, in which at the stage of determining the optimal bandwidth, the optimum bandwidth corresponding to the distribution coefficient, determined by reference to the table determine the optimal bandwidth, in which you set the dependency relationship between the distribution coefficient and the optimum bandwidth.

9. The video processing method of claim 8, in which at the stage of determining the optimal bandwidth when many tables determine the optimal bandwidth in accordance with the size of the image and the target value of the error information, select table determine the optimal bandwidth, which corresponds to the size of the area of separation and a given target value of the error information, and the optimum bandwidth corresponding to the distribution coefficient, determined by reference to the selected table to determine the optimal bandwidth.

10. The video processing method according to claim 1, in which the step of calculating the distribution ratio is performed through the your dividing the value of the error information, which is received in a state very close to the condition in which there is restriction bands, the displayed value of the error information.

11. The rendering of containing phases in which: determines a first bandwidth applied to

areas of separation, which are set on the target image processing, forming the video image, so as to be divided into sections destination image processing;

calculate a first array of filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, by using the first bandwidth;

shape data of the filtered image by the exposure data of the target image processing filtering process using the first array of filter coefficients;

output for each area of the division is information about the error between the target image processing and data filtered image, and calculate the distribution coefficient used for determining an optimum bandwidth, based on the extracted values;

determine for each area of the separating optimum bandwidth corresponding to the distribution coefficient;

calculate for each region of the separation pattern optimalen the x filter coefficients for implementing frequency characteristics,
corresponding to the limitation of bandwidth, using a specific optimal bandwidth;

form data optimum filtered divided image of each region dividing by the exposure of the image data of each region separating the filtering process using an array of optimal filter coefficients; and

synthesize data optimum filtered divided image of each area of separation.

12. The video processing method according to claim 11, in which the step of determining the first bandwidth, the first bandwidth is determined based on the size of each field division.

13. The video processing method according to claim 11, further comprising stages, in which: compare for each area of the separation of certain optimal bandwidth with optimal bandwidth of the peripheral area of separation around this area separation; and adjust a certain optimal bandwidth based on the comparison result.

14. The video processing method according to claim 11, further comprising stages, which determine whether the motion image data of each region, division, using the image data of the frame before or after the frame area or image data to the wood before and after the frame region; and correct certain optimal bandwidth each area of the division for which it is determined that the image data area splitting have movement.

15. The video processing method according to 14, in which the step of determining whether the moving image data of each area separation is performed by evaluating the variations of the pixel values from each frame to the given frame.

16. The video processing method according to 14, in which: the step of determining whether the moving image data of each area separation is performed by determining whether the movement data of the image region separation, and is characterized by high-frequency component; and a phase adjusting certain optimal bandwidth is performed by adjusting a certain optimal bandwidth each area of separation, which determined that these images have motion, and characterized by high-frequency component.

17. The video processing method according to clause 16, in which: the step of determining whether the moving image data of each region, division, and characterized by high-frequency component is performed by stages, in which: determines whether the value that represents the attribute data of image is of Azania the area of separation, these images are characterized by high-frequency component; and assess the variation in the number of areas of separation, which shows that the image data are characterized by high-frequency component, from each frame to the given frame.

18. The video processing method according to claim 11, in which at the stage of determining the optimal bandwidth, the optimum bandwidth corresponding to the distribution coefficient, determined by reference to the table determine the optimal bandwidth, in which you set the dependency relationship between the distribution coefficient and the optimum bandwidth.

19. The rendering of on p, in which at the stage of determining the optimal bandwidth when many tables determine the optimal bandwidth in accordance with the size of the image and the target value of the error information, select table determine the optimal bandwidth, which corresponds to the size of the area of separation and a given target value of the error information, and the optimum bandwidth corresponding to the distribution coefficient, determined by reference to the selected table to determine the optimal width of p is band width.

20. The video processing method according to claim 11, in which the step of calculating the distribution ratio is performed by dividing the value of the error information, which is received in a state very close to the condition in which there is limitation of bandwidth, the displayed value of the error information.

21. The video processing device, comprising:

unit for dividing the target image processing, which forms a video image into many regions of separation;

the unit for determining the first bandwidth applied to the areas of separation;

unit for calculating a first array of filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, by using the first bandwidth;

unit to generate the data filtered divided image for each area of the division, by exposure of the image data of each region separating the filtering process using the first array of filter coefficients;

block to derive for each area of the separation values for the error information between the image data of each area separation data and the filtered divided image, and calculate the distribution factor used to determine the optimal width is not the bandwidth,
on the basis of the extracted values;

unit for determining for each area of the separating optimum bandwidth corresponding to the coefficient of distribution;

unit for calculating for each area of the separation of the array of optimal filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, using a specific optimal bandwidth;

the block data generating optimum filtered divided image of each region dividing by the exposure of the image data of each region separating the filtering process using an array of optimal filter coefficients; and

block for the synthesis of data the optimum filtered divided image of each area of separation.

22. The video processing device according to item 21, further comprising:

block for comparison for each area of the separation of certain optimal bandwidth with optimal bandwidth of the peripheral area of separation around this area separation; and

unit to adjust a specific optimal bandwidth based on the comparison result.

23. The video processing device according to item 21, further comprising: a unit for determining whether motion is image data of each region, division, using the image data of the frame before or after the frame area or image data frames before and after the frame region; and a unit for correcting certain optimal bandwidth each area of the division for which it is determined that the image data area splitting have movement.

24. The video processing device according to item 23, in which: a unit for determining whether the moving image data of each region, division, determines whether the moving image data area splitting and are characterized by whether they are high-frequency component; and a unit for correcting certain optimal bandwidth adjusts certain optimal bandwidth each area of separation, which determined that these images have movement and are characterized by high-frequency component.

25. The video processing device, comprising:

the unit for determining the first bandwidth based on the size of the area of separation for separation, which is set on the target image processing, forming the video image, so as to be divided into sections destination image processing;

unit for calculating a first array of filter coefficients for implementing frequency characteristics corresponding to the restriction on the wasps,
using the first bandwidth;

unit to generate the data of the filtered image by the exposure data of the target image processing filtering process using the first array of filter coefficients;

block to derive for each area of the separation values of the errors between the data of the target image processing and data filtered image and calculate the distribution coefficient used for determining an optimum bandwidth, based on the extracted values;

unit for determining for each area of the separating optimum bandwidth corresponding to the coefficient of distribution;

unit for calculating for each area of the separation of the array of optimal filter coefficients for implementing frequency characteristics corresponding to the limitation of bandwidth, using a specific optimal bandwidth;

the block data generating optimum filtered divided image of each region dividing by the exposure of the image data of each region separating the filtering process using an array of optimal filter coefficients; and

block for the synthesis of optimal data filtered razdelno what about the image of each area of separation.

26. The video processing device according A.25, further comprising: a unit for comparison for each area of the separation of certain optimal bandwidth with optimal bandwidth of the peripheral area of separation around this area separation; and a unit for correcting certain optimal bandwidth based on the comparison result.

27. The video processing device according A.25, further comprising: a unit for determining whether the moving image data of each region, division, using the image data of the frame before or after the frame area or image data frames before and after the frame region; and a unit for correcting certain optimal bandwidth each area of the division for which it is determined that the image data area splitting have movement.

28. The video processing device according to item 27, in which: a unit for determining whether the moving image data of each region, division, determines whether the moving image data area splitting and are characterized by whether they are high-frequency component; and a unit for correcting certain optimal bandwidth adjusts certain optimal bandwidth each area of the division for which the determination is prohibited, that its image data have movement and are characterized by high-frequency component.

29. The computer-readable storage medium that stores the program and handles, with which the computer executes a process for implementing the video processing method according to claim 1.

30. The computer-readable storage medium that stores the program and handles, with which the computer executes a process for implementing the video processing method according to claim 11.

**Same patents:**

FIELD: information technologies.

SUBSTANCE: parametres of coding, such as tables of end of block (EOB) shift and selection of dictionary of VLC-coding may be stored as internal conditions instead of their sending with coded data of series of serial macroblocks of video. Records of tables may periodically be updated on the basis of statistics collected in process of coding stage. Table of special EOB shift may adapt position of symbol of special EOB in set of symbols to probability of significant coefficients with absolute value, more than 1, for condition of coding, such as cycle of coding. Units of colour signal may be coded independently on blocks of brightness signal with the help of separate table of EOB shift tables, shift of special EOB and selection of VLC-coding dictionary.

EFFECT: increased efficiency of adaptive coding with alternating length.

54 cl, 11 dwg

FIELD: information technologies.

SUBSTANCE: fragments may be aligned with cycle so that beginning of working data of each of fragments practically matches beginning of one of cycles. Certain cycles may be controlled with the help of vector mode to scan for previously determined position within the framework of block prior to displacement to another block. Therefore, number of cycles may be reduced, cutting number of fragments and associatively related volume of service information. CAF may be entropically coded independently on each other, so that each of fragments could be easily accessible and decoded without waiting for decoding of other fragments, which provides for parallel decoding and simultaneous processing of these fragments.

EFFECT: increased efficiency of video coding based on cycles, coding of coefficients of blocks of FGS-video data and syntactic elements, reduction of fragments number and volume of service information.

68 cl, 14 dwg, 8 tbl

FIELD: information technologies.

SUBSTANCE: system and method are proposed to indicate points of switching from lower level to higher one at the level of file format for efficient switching of scalable flows in flow servers and at local reproduction of files. Also the present invention proposes system and method for indication of switching points from lower level to higher in bit video flow, for instance to provide for intellectual forwarding of scalable levels in network elements, capable of recognition of media data, or computationally scalable decoding in receivers of flow.

EFFECT: provision of simple and efficient switching from lower level to higher one for adaptation of scalable flow without necessity in detailed selection and analysis of bit video flows by flow server, and thus reduction of computational and implementation resources.

16 cl, 6 dwg, 2 tbl

FIELD: information technologies.

SUBSTANCE: video data of improvement level is included into component of network abstraction level (NAL); one or several syntactic elements are included into NAL to indicate, whether NAL component includes video data of improvement level, and one or several syntactic elements to indicate at least one of the following: whether video data of improvement level include video data with intraframe coding in NAL component; whether NAL component includes parametre of sequence, set of image parametres, layer of reference image or section of data of reference image layer; and number of nonzero coefficients with value exceeding one, in units of intraframe coding in video data of improvement level.

EFFECT: realisation of expansion of versions of International Telecommunication Union standard for efficient scalable video coding.

49 cl, 16 dwg, 18 tbl

FIELD: information technology.

SUBSTANCE: scalable video signal encoder includes an encoder for encoding a block in an enhancement layer of a picture by applying a same weighting parametre to an enhancement layer reference picture as that applied to a lower layer reference picture used for encoding a block in a lower layer of the picture. The block in the enhancement layer corresponds to the block in the lower layer, and the enhancement layer reference picture corresponds to the lower layer reference picture.

EFFECT: high efficiency of weighted prediction for scalable coding and decoding a video signal with possibility of storing different sets of weighting parametres for the same reference picture in the enhancement layer.

31 cl, 6 dwg

FIELD: information technologies.

SUBSTANCE: device and method are proposed to process multimedia data, such as video data, audio data, or video and audio data for coding, using certain classification of content. Processing of multimedia data includes determination of multimedia data complexity, classification of multimedia data on the basis of certain complexity, and determination of transfer speed in bits for coding of multimedia data on the basis of their classification. Complexity may include a component of spatial complexity and component of time complexity of multimedia data. Multimedia data is classified, using classifications of content, which are based on value of visual quality for viewing of multimedia data, using spatial complexity, time complexity or both spatial and time complexity.

EFFECT: development of improved method of images classification.

111 cl, 12 dwg

FIELD: information technologies.

SUBSTANCE: method for decoding of compressed video sequence, at the same time image frames are introduced into buffer memory related to decoding. Video sequence includes indication related to at least one gap in numbering of image frames, besides this indication is decoded from video sequence. Further, in response to this indication, buffer memory is configured so that it provides for number of images frames corresponding to gap in numbering of image frames, and images frames in buffer memory are used in process of decoding. Preferentially, specified indication informs about the fact that at least one gap in numbering of image frames in video sequence is deliberate, and specified number of image frames is used in buffer memory instead of image frames, which are not available in decoder.

EFFECT: provision of the possibility for decoder to account for image frames, which were deleted deliberately by coder.

31 cl, 14 dwg

FIELD: information technology.

SUBSTANCE: subsets are determined (step 29), each containing one or more coding units, where at least one image puts at least one coding unit into two or more subsets, the list of requirements (LOR) is established (step 30) containing at least one element associated with each subset. Significance values are use in order to select quality increments for generating an allowable code stream which satisfies the LOR for subsets (steps 34, 36). Quality increments can be selected so as to attain high quality for different subsets depending on size requirements in the LOR. For certain requirements, the code stream will exhibit an approximately constant quality of the reconstructed image. Quality increments can be selected so as to achieve small sizes of a compressed image for different subsets depending on quality requirements in the LOR.

EFFECT: high quality of the reconstructed image.

27 cl, 7 dwg

FIELD: information technology.

SUBSTANCE: coding device has definition apparatus for determining image area data meant for processing in order to counter reconstruction implied by granular noise arising in image data coded based on said image data and apparatus for countering reconstruction, designed for processing in order to counter reconstruction for image area data, defined using definition apparatus when coding image data, where when the said image data are coded in data unit data modules, the said definition apparatus determines unit data which form the said image data as the said image area data, and apparatus for countering reconstruction forcibly sets the orthogonal transformation coefficient to zero, which becomes equal to zero when quantisation is carried out using the said unit data, among orthogonal transformation coefficients of unit data defined using the said definition apparatus.

EFFECT: improved quality of the decoded image.

14 cl, 18 dwg

FIELD: information technologies.

SUBSTANCE: method includes the following: operation of coding for formation of coded images in coder, and also operation of specified coded images transfer in decoder in the form of transmission units, operation of buffering for buffering of transmission units sent to decoder, in buffer and operation of decoding for decoding of coded images with production of decoded images. Buffer size is specified by means of determining overall size of at least two transmission units and setting maximum size of buffer on the basis of this overall size.

EFFECT: improved efficiency of coded images buffering.

22 cl, 16 dwg

FIELD: electricity.

SUBSTANCE: filter includes the first inductance coil to the outputs of which there connected is the second and the third inductance coil, the second outputs of the second and third coils are connected to common bus, as well as two capacitors of variable capacity, the first outputs of capacitors are connected to common bus; besides filter includes the fourth inductance coil connected to input potential terminal of filter and to the second output of the first capacitor, fifth inductance coil connected to output terminal of filter and to the second output of the second capacitor, the sixth inductance coil connected to the second output of the first capacitor and to the first output of the second inductance coil, the seventh inductance coil connected to the second output of the second capacitor and to the first output of the third inductance coil.

EFFECT: improving amplitude-versus-frequency characteristics at large values of tuning range.

4 dwg

FIELD: electricity.

SUBSTANCE: rejection filter including two inductance coils which are in-series connected; the first output of the first one of them is connected to input potential terminal of filter; the first output of the second inductance coil is connected to output potential terminal of filter; to the second outputs of these inductance coils there connected is the first and the second capacitors, as well as the third inductance coil the second output of which is connected through the third capacitor to common bus; besides filter includes the fourth and the fifth inductance coils; at that, the fourth inductance coil is connected between input potential filter terminal and the second output of the first capacitor; the fifth inductance coil is connected between output potential filter terminal and the second output of the second capacitor.

EFFECT: improving manufacturability.

2 dwg

FIELD: physics.

SUBSTANCE: band-pass LC-filter with controlled transmission bandwidth which consists of first and second inductor coils whose first leads are connected to a first capacitor and a third inductor coil whose second leads are connected to a common bus, also contains fourth and fifth inductor coils, second and third capacitors and two varicaps, where the fourth inductor coil is connected to the input potential terminal of the filter and the second lead of this coil is connected to first leads of the second capacitor and the first varicap. The fifth inductor coil is connected to the output potential terminal of the filter and the second lead of this coil is connected to first leads of the third capacitor and the second varicap. Second leads of the varicaps, second and third capacitors are connected to the common bus.

EFFECT: control of transmission bandwidth during operation.

2 dwg

FIELD: physics.

SUBSTANCE: variable LC-band pass filter consists of n cascaded L-shaped half sections of high-pass filters, each having capacitors in a series circuit. The first lead of the capacitor is the first input of the half section and the second lead is the output which is connected to an inductor coil whose second output is connected to another capacitor. The input of the first half section is connected to the input potential terminal of the filter. The second leads of the capacitors which are connected in parallel circuits of the half sections are connected to a common bus. The output of the last half section is cascaded with m L-shaped half sections of low-pass filters, each consisting of an inductor coil and in a series circuit and a capacitor connected in parallel. The first leads of this inductor coil and the capacitor are the input of the half section and their second leads are the output which is connected to another capacitor. The output of the last half section is connected to the output potential terminal of the filter and the second leads of capacitors in the parallel circuit of the last ℓ half sections low-pass filters are connected to the common bus. k<n and ℓ<m are any integers. The invention is distinguished by that the filter has four groups of switches. The first group consists of n-k switches whose first leads are connected to input leads of half sections of the low-pass filter, starting with the k+1-th to the n-th half section respectively. Second leads of these switches are connected to the output of the high-pass filter. The second group consists of n-k switches whose first leads are connected to second leads of n-k capacitors connected in the parallel circuits of high-pass filters starting with the k+1-th half section to the n-th half section respectively. The second leads of these switches are connected to the common bus. The third group has m-ℓ switches whose first leads are connected to inputs of half sections of low-pass filters starting with the ℓ+1-th to the m-th, respectively. The first leads of these switches are connected to the output of the m-th half section. The fourth group consists of m-ℓ switches whose first leads are connected to second leads of m-ℓ capacitors connected in the parallel circuits of low-pass filters starting with the ℓ+1-th half section to the m-th half section, respectively. The second leads of these switches are connected to the common bus.

EFFECT: simplification.

1 dwg

FIELD: radio engineering.

SUBSTANCE: tunable band filter comprises N+1 inductances, which are connected serially, according capacitance and inductance, the second outputs of which are connected to common bus, to each of neighbouring inductances connection points. This chain creates band filter, which consists of N cascade connected links, input of specified band filter via the first inductance is connected to input potential terminal of device, output of this band filter is connected via the second inductance to outlet potential terminal of device, points of connection of neighbouring inductances, which are included into longitudinal branches of specified N link band filter, and also points for connection of the first and second inductances with this band filter produce N+2 intermediate outputs of device and are connected accordingly to N+2 inputs of commutator additionally introduced in device, M outputs of which are connected to capacitance box, comprising M capacitors, the first outputs of which are connected to M outputs of commutator, and the second outputs are connected to common bus.

EFFECT: expansion of functional resources.

1 dwg

FIELD: radio engineering.

SUBSTANCE: invention is intended for fine tuning of multilink band filters (hereinafter referred to as filters) of radio electronic devices of microwave and in compliance with the specified requirements. Result is achieved by tuning of filter by specific points of frequency dependence of transfer coefficient (maximum point (7) - in regulation of links, point of maximum slope (6) - in tuning of resonators) at partial connections of filter into measuring track of device, which make it possible to simultaneously control both frequency of resonator tuning and links size. At the same time calculation is simplified for variation of resonator tuning frequency and links size. Method makes it possible to tune two and more filters with the same parametres. It is not required to make any absorbers, accessories for shorting, devices for energy drain.

EFFECT: lower labour intensiveness of tuning process.

3 dwg, 2 ex

FIELD: physics.

SUBSTANCE: invention refers to tuneable pass-band filters (PBF) of receivers and transmitters. In input and output PBF contains, two serial communications, i.e. inductive (1) and transformer (4), and between PBF resonant circuits transformer coupling communications (10) designed as structural component with windings on one frame. Note that communication windings of non-adjacent resonant circuits are located on both sides relative to winding of centre resonant circuit.

EFFECT: maintenance of constant relative pass-band within whole operating frequency range.

2 dwg

FIELD: radio engineering; adjusting frequency-dependent filters.

SUBSTANCE: proposed regulating device designed to adjust frequency response of filter has time-constant transducer that function to measure state of filter charge and to withdraw measurement results from mentioned time-constant transducer in the form of first voltage; converter used to digitize first voltage and to withdraw second voltage obtained as result of first-voltage conversion by mentioned converter; array of adjusting components; and selector using second voltage for selecting at least one adjusting component out of array of adjusting components; state of circuit charge can be adjusted in proposed device by means of at least one chosen adjusting component; mentioned time-constant transducer of proposed device has in addition change-over controller that functions to control at least one charge switch; closure of the latter switches filter over to state of charge.

EFFECT: reduced admissible differences between measured and desired values of frequency-dependent filter components.

14 cl, 6 dwg

FIELD: radio engineering, possible use in radiolocation stations working with two multi-frequency generators.

SUBSTANCE: directional filter contains two low-and-high-pass filters, a T-joint and two two-channel commutators. Channels of commutator rotors are coupled with shoulders of T-joint, which amount to λ/4 or λ/2 respectively with consideration of rotor channel lengths, where λ - wave length on shoulders of the T-joint.

EFFECT: ensured transmission of two UHF signals of different frequencies sequentially or simultaneously, and of multi-frequency signals - sequentially to one antenna from two transmitters.

4 dwg

FIELD: information technology.

SUBSTANCE: first band pass (BP) is determined based on initial image data; a matrix of filter coefficients (FC) is calculated to obtain frequency characteristics corresponding to limitation of frequency band (FB) using the first BP; data of the first filtered image are generated by filtering data of the initial image using the matrix of first FC; an estimate value of the objective image quality of data of the first filtered image is obtained and the distribution coefficient (DC) is calculated, which is used to determine the optimum BP based on the estimate value of objective image quality; the optimum BP corresponding to the calculated DC is determined using a table in which the corresponding relationship between DC and optimum BP is defined; a matrix of optimum FC is calculated to obtain frequency characteristics corresponding to limitation of FB using the optimum BP; and data of the optimally filtered image is generated by filtering data of the initial image using the matrix of optimum FC.

EFFECT: adaptive image filtering process for providing high-quality image.

3 cl, 11 dwg