The way the signal adaptive filtering, signal adaptive filter and machine-readable media (options) for storing programs

 

The invention relates to computing and can be used for image processing. The technical result is to increase the quality of the image. Methods based on what is produced by the information block to reduce the effects of blocking and edging for noise reduction outline of the coefficients of predetermined pixels in the boundary regions of the data block. When this is done adaptively filtering the image data through inverse quantization and inverse discrete cosine transformation in accordance with the agreed information block and the information of the border. 3 S. and 14 C.p. f-crystals, 7 Il.

The invention relates to filter the data, in particular to the way the signal adaptive filtering for reducing blocking effect and noise trim, signal adaptive filter and machine-readable media for storing the program.

The prior art In General, the standards-based image coding, such as MPEG standards of the International organization for standardization (ISO) and N. 263 recommended by the International telecommunication Union (ITU), accept based on units of assessment is but using block-based coding can cause the well-known blocking effect. The usual effect of blocking is the noise of the grid in a homogeneous region in which adjacent pixels have the same pixel values. Another effect of blocking the stair is noise, which has the shape of a ladder and is produced along the edges of the image. In addition, the noise fringing is a common Gibbscam phenomenon resulting from truncation, when the DCT coefficients quanthouse so to compress the image as much as possible.

The noise of the grid shows the traces based on the units of the process on the edges between the blocks, when compressed data is displayed on the screen after restore. Thus, it is possible to identify edges between blocks. In addition, staircase noise has the form of a ladder at the edges of the image, so you may notice uneven edge on the image. You also notice the overlapping images with a preset interval because of the noise of the border.

In order to reduce the effect of blocking and noise edging provided when using block-based coding, suggested several ways. In accordance with the encoding standard N. 261, for snii (LPF) 3x3 [Video Codec for audiovisual Services at RH kbit/s. CCITT Recomendation H. 261, December 14, 1990 (Video codec for audiovisual services at Px62 Kbit/s - CCITT recommendation H. 261, 14 December 1990)]. In addition, a simple edge filter in the feedback circuit was proposed in order to reduce the effect of blocking and mosquito noise [G. Bjorntegaard, "A simple Edge Loop Filter to Reduce Blocking and Mosquito Noise", ISO/IEC JTC/Sc29/WG11 MPEG96/0617, January, 1996 (Simple edge filter in the feedback circuit to reduce the effect of blocking and mosquito noise]. Edge filter in the feedback circuit produces linear values of two pixels adjacent to the edge of the block, and replaces the values of the two pixels of the linearized values. This edge filter in the feedback circuit can reduce the blocking effect, but not the noise of the border. For noise reduction trim was proposed nonlinear filter using a binary index [Y. Itoh, "Detail Preserving Nonlinear Filter using Binary Index", ISO/IEC JTC/SC29/WG11 MPEG95/0357, November, 1995 (Protecting details of the nonlinear filter using a binary index)]. However, the non-linear filter cannot reduce the blocking effect.

Disclosure of the invention To solve the above technical problems the present invention is the provision of a method of signal adaptive filtering to reduce the effect of BC is aemy media.

In accordance with one aspect of the present invention, a method of signal adaptive filtering can reduce the effect of blocking and noise outline of the image data when the frame is formed of blocks of a predetermined size containing the following: (a) elaboration of the information block to reduce the effect of blocking and information edging for noise reduction outline of the coefficients of predetermined pixels of the upper and left boundary regions of the data block when the frame obtained by decomposition of an image data stream of binary digits for inverse quantization, an inner frame; and (b) adaptive filtering of the image data, passing through inverse quantization and inverse discrete cosine transformation, in accordance with the agreed information block and the information of the border.

Preferably, the operation (a) further includes the setting information block and the information of the outline of the previous frame corresponding to the motion vector, as the information block and the information of the outline of the current frame, if the frame is an inner frame, and installation information edging to "1" quantized current block, and information block and the information of the outline is determined in accordance with the coefficients of the pixel And located in the upper left corner of this block, the pixel At the right of the pixel a and the pixel below the pixel A.

Preferably, the information block consists of a horizontal information block and the information of the vertical block, and the horizontal information block is set to "1", which means the requirement of filtering for the image data, when only the coefficient of the pixel is not equal to "0" or any coefficient of the pixels of the left boundary region of the block is not equal to "0", and the vertical information block is set to "1", which means the requirement of filtering for the image data, when only the coefficient of the pixel is not equal to "0" or any coefficient of the pixels of the upper boundary region of the block is not equal to "0", and information edging is set to "1", which means the requirement of filtering for the image data, when any coefficient of the pixels that are different from the pixels In the block is not equal to "0".

Preferably, the information block consists of a horizontal information block and the information of the vertical lock for the image data, when all the coefficients of the pixels In the block is not equal to "0" or any coefficient of the pixels of the left boundary region of the block is not equal to "0", and the vertical information block is set to "1", which means the requirement of filtering for the image data, when all the coefficients of the pixels a, b, C is not equal to "0" or any coefficient of the pixels of the upper boundary region of the block is not equal to "0" and the information edging is set to "1", which means the requirement of filtering for the image data, when any coefficient of the pixels, different from the pixels In the block is not equal to "0".

Preferably, in order to reduce the blocking effect, in operation (b) horizontal (or vertical) filtering is performed using a weighted filter having a predetermined weighted value when the information of the horizontal (or vertical) block to block is "1" and information edging "0", and when the information of the horizontal (or vertical) block to block is not equal to "1" or information edging is not equal to "0", compares the absolute value of the difference between adjacent pixels and the value of Q used as the dividend for the quantization unit, and then filter with in advance for what I am, proposed signal adaptive filter, can reduce the effect of blocking and noise outline of image data when a frame is composed of blocks of a predetermined size, comprising: a unit test mode flag to check the flag to determine whether the frame is an inner frame and an intermediate frame when the image data in the stream of binary bits are arranged to inverse quantization; generator information about intra-frame filtering to generate the information block for reducing blocking effect and to generate information edging for noise reduction trim, from the coefficients of predetermined pixels of the upper and left boundary regions of the data block, when the frame is defined by a block check mode flag as the inner frame; a generator of information about the interframe filter for installation information block and the information of the outline of the previous frame in accordance with the motion vector information block and the information of the outline of the current frame, if the frame is an intermediate frame, and installation information edging to "1" if there is a residual signal of the inverse quantized current block; and an adaptive filter for adaptive healrather, in accordance with the information block and the information of the outline developed by the generator information about intra-frame filtering and generator information interframe filtering.

The invention may be embodied in digital General-purpose computer that executes the programs used in the computer media, including, but not limited to such data carriers, magnetic data carriers (for example, ROM, floppy disks, hard disks, etc), optically readable media (e.g. CD-ROM, DVD, etc) and bearing vibrations (e.g., transmission via Internet). Therefore, the present invention can be embodied as used in the computer media.

In accordance with another aspect of the present invention, a machine-readable medium having embodied thereon a computer program for signal adaptive filtering can reduce the effect of blocking and noise outline of the image data when the frame is formed of blocks of a predetermined size, where the signal adaptive filtering includes the following operations: (a) elaboration of the information block to reduce the effect of blocking and information edging to reduce noise Okant is obtained by decomposition of an image data stream of binary digits for inverse quantization, an internal frame; (b) installation information block and the information of the outline of the previous frame in accordance with the motion vector information block and the information of the outline of the current frame, if the frame is an inner frame, and setting information edging to "1" that represents the request filtering for the image data, if there is a residual signal of the inverse quantized current block, and (C) adaptive filtering of the image data passing through is based on units of inverse quantization and inverse discrete cosine transformation, in accordance with the agreed information block and the information of the border.

Also proposed machine-readable medium having embodied thereon a computer program for a method of filtering noise fringing caused by decoding compressed on the basis of blocks of image data, where the method of filtering noise outline contains the following operations: (a) performing a gradient operation on the block subjected to inverse quantization and inverse discrete cosine transform, using a predetermined one-dimensional horizontal and vertical gradient operators; (b) vertigo difference between treated gradient operator value of one pixel and the value of the adjacent pixel, and the values of Q used as the dividend for the quantization unit; and (C) performing filtering by applying a predetermined filter window to the generated binary edge map.

A brief description of the drawings Fig. 1 is a block diagram of a signal adaptive filter for reducing blocking effect and noise edging in accordance with the present invention.

Fig. 2 is a block diagram of the algorithm, illustrating how the signal adaptive filtering in accordance with the preferred implementation of the invention.

Fig.3 shows the inverse quantized block with 8x8 pixels.

Fig. 4 is a block diagram of the algorithm, illustrating the operation generation information used to filter the internal frame.

Fig. 5 is a flowchart of an algorithm illustrating the operation generation information used for filtering the intermediate frame.

Fig.6 shows the arrangement of pixels adjacent to the edge of the frame, to illustrate filtering for reducing blocking effect.

Fig.7 shows the location of pixels to be processed in the current block.

The best implementation of the invention In Fig.1 signal adaptive filter to reduce the effect of mode, the generator 130 information about intra-frame filtering, the generator 140 information interframe filtering and block adaptive filter 150. When the image data in the stream of binary bits are arranged for the inverse quantization unit 120 checks the mode flag checks whether the frame is an inner frame and an intermediate frame. The generator 130 information about inside-HR filter produces information block for reducing blocking effect and the outline of the coefficients of predetermined pixels of the upper and left boundary regions of the data block when the frame defined by the block 120 checks the mode flag as an internal frame. When the unit 120 checks the mode flag identifies the frame as an intermediate frame generator 140 information interframe filtering produces information block and the information of the outline of the previous frame in accordance with the motion vector information block and the information of the outline of the current frame. If the residual signal of the inverse quantized current block exists, information edging is set to "1". Block 150 adaptive filter adaptive filters the image data which is passed through the inverse quantizer 100 (K- the education and information edging, produced by generator 130 information about intra-frame filtering and generator 140 information interframe filtering.

Will now be described by way of the signal adaptive filtering in accordance with a preferred implementation of the present invention. Fig.2 is a block diagram of the algorithm, illustrating how the signal adaptive filtering in accordance with the present invention. The image data in the stream of binary bits encoded by the encoder are decoded by the decoder for playback. To end this, the data in the stream of digits are displayed and then the inverse quanthouse inverse quantizer 100 (operation 200). Here, the image data consist of multiple frames, and each frame consists of multiple blocks. Fig.3 shows the inverse quantized block with 8x8 pixels, which form a frame.

Before filtering the data frame with the inverse discrete cosine transform (matchcode ID) checks the flag for the issue of determining whether a frame is an inner frame and an intermediate frame (operation 210). If the frame is an inner frame (operation 220), produces information that is used for intraframe filter (operation 230). If the frame is the eat data frame, they went through the matchcode ID 110, adaptive filters in accordance with the agreed information filter, thereby eliminating the effect of blocking and noise trim (operation 250).

Fig. 4 is a block diagram of the algorithm in detail illustrating the operation generation information used to filter the internal frame. As shown in Fig.4, if the frame is determined by the block 120 checks the mode flag as an internal frame, it checks the ratio of the pixel And Fig.3 (operation 400). If the ratio of the pixel is not equal to "0", the horizontal block (JGB), and the vertical block (MCB) are set to "1" (operation 410). If any coefficient of the pixels (8 pixels, including pixels a and b, belonging to the upper edge area of the 300 block shown in Fig. 3, is not equal to "0" (operation 420), WBI is set to "1" (operation 430). Otherwise WBI is set to "0" (operation 440). Also, if any coefficient of the pixels (8 pixels, including pixels a and C), belonging to the left boundary region 310 of the block shown in Fig.3, is not equal to "0" (operation 450), JGB is set to "1" (operation 460). Otherwise JGB is set to "0" (operation 470).

After JGB and WBI installed, produces information the pixels And, In block shown in Fig.3, is not equal to "0" (operation 480), IO is set to "1" (operation 490). Otherwise, IO is set to "0" (operation 470). Here, JGB and WBI is set to "1" only when the ratio of the pixel is not equal to "0" (operation 400). However, by setting up and WBI to "1" even if all the coefficients of the pixels a, b and C is not equal to "0", the favorable effect to some extent can be obtained when the signal adaptive filtering is performed later.

Fig. 5 is a flowchart of an algorithm illustrating the operation generation information used for filtering the intermediate frame. If the frame is determined by the block 120 checks the mode flag as an intermediate frame, JGB, WBI and IO internal frame is converted into JGB, WBI and IO intermediate frame in accordance with the motion vector (operation 500). Also, if the residual signal exists after motion compensation (operation 510), IO is updated (operation 520).

When the information block and the information edging to filter produced as described above, filtering is adaptive in accordance with the information. First will be described a method of filtering to reduce the effect of blocking. Filtering to reduce the effect of th is t clarified horizontal filtering. Fig.6 shows the arrangement of pixels adjacent to the edge of the frame, to illustrate filtering for reducing blocking effect. Is performed to determine whether "0" JGB and IO blocks I and J in Fig.6. If JGB and IO blocks I and J in Fig.6 is equal to "0" weighted filtering is performed for the pixels P1, P2, P3, P4, P5 and P6 in Fig. 6 using 7-outlet(1, 1, 1, 2, 1, 1, 1) low pass filter (LPF).

If JGB or IO blocks I and J in Fig.6, is not equal to "0", the filtering is performed for the pixels P2, P3, P4 and P5 using the following algorithm: P4=P4; If (ABS(d)(Q) { P4=P4-(d/2); P3=P3+(d/2); d=P5-P4; if (ABS(d)(Q); P5=P5-(d/4);
d=P3-P2;
if (ABS(d)(Q); P2-P2+(d/4);
}
else {
if (ABS(d/2)(2Q)
If (d>0) {
P4=P4-(Q-ABS(d/2));
P3=P3+(Q-ABS(d/2));
}
else {
P4=P4+(Q-ABS(d/2));
P3=P3+(Q-ABS(d/2));
}
d=P5-P4;
if (ABS(d)(Q); P5=P5-(d/4);
d=P3-P2;
if (ABS(d)(Q); P2=P2-(d/4);
}
}
In the above algorithm, ABS represents the absolute value, and Q is the dividend used when quanthouse blocks that make up the frame.

More, the absolute value (ABS(d)) difference (a) between the pixels P4 and P3 is equal to or less than Q, current is the total pixel value of the pixel P3 is set by adding d/2 in the current pixel value. In addition, the absolute value (ABS(d)) of the difference (d) between the pixels P5 and P4 is equal to or less than Q, the current pixel value of the pixel is set by subtracting d/4 from the current pixel value. In addition, if the absolute value (ABS(d)) of the difference (d) between the pixels P3 and P2 is equal to or less than Q, the current pixel value of the pixel P2 is set by subtracting d/4 from the current pixel value. Similarly to the above, the pixel values of the pixels P2, P3, P4 and P5 are set in accordance with the algorithm, other than the above, which is obvious to the expert, so that its explanation is omitted. In addition, vertical filtering is performed according to the principle similar to the principle of horizontal filtering.

Next will be described a method of filtering for noise reduction trim. First, it is checked generated information edging. If the trim is set to "1", the filtering occurs. Otherwise no filtering is performed. For the end of the determined edge pixels of the block that had been subjected to inverse quantization and matchcode ID. To determine the edge pixels in the blocks which have been subjected to inverse quantization and matchcode ID, gradient is the e value of the difference between the one treated gradient operator pixel value and the value of the neighboring pixel, and the value of Q used as dividend during the quantization unit, used to generate a binary edge map representing the edges of each pixel. Here the block is 8x8 pixels, and the size of the binary maps of the edges appears to be a two-dimensional matrix region [10][10], as shown in Fig.7.

To generate a binary edge map are detected vertical edges and the detected horizontal edges. Algorithms for the detection of vertical edges and horizontal edges are:
/Detection of vertical edges"/
A1=ABS(PrtImage[0]-Prtlmage[1]);
A2=ABS(Prtlmage[0]-Prtlmage[-1]);
if

Edge[m][n]=1; /edge/
else {/detect horizontal edges"/
And'1=ABS(Prtlmage[0]-Prtlmage[width]);
And'2=ABS(Prtlmage[0]-Prtlmage[-width]);
if

Edge [m][n]=1; /edge/
}
To detect vertical edges calculates the absolute value (A1) of the results of applying the gradient operator to the pixel (Prtlmage[0]), this value is determined whether the pixel is on the edge of the block or not, and right pixel (Prtlmage[1]) of the pixel (Prtlmage[0] ). Then calculate the absolute value (A2) of the results of applying the gradient operator to the pixel (Prtlmage[0]) and the left who is depending on a Boolean value, received after the absolute values of A1 and A2 are compared with a preset threshold value Th, and then the above process is performed for all pixels of the block. Detection of vertical edges is performed in accordance with a logical formulaIf this logical formula is true, the pixel is determined as a vertical edge. Otherwise, the pixel is determined as the non-vertical edge.

Detecting horizontal edges is performed in accordance with the same principle of detecting horizontal edges. First, calculate the absolute value (A'1) difference between treated gradient operator the results of the pixel (Prtlmage[0]), which is determining whether the pixel is on the edge of the block or not, and bottom pixel (Prtlmage[width]) of the pixel (Prtlmage[0]). Then calculate the absolute value (A'2) the difference between the treated gradient operator the results of the pixel (Prtlmage[0]) and the upper pixel (Prtlmaget-width]) pixel (Prtlmage[0] ). Thus, the determination of whether the pixel is on the edge, performed according to the logical values obtained after the absolute value of A'1 and a'2 are compared with zarabianie horizontal edges is performed in accordance with a logical formula. If this logical formula is true, the pixel is determined as a horizontal edge. Otherwise, this pixel is determined as a non-horizontal edge. Here, "& & " represents a Logical And, and "" represents a Logical OR.

Further filtering is performed by applying a predetermined filter window to the generated binary map edges. Filtering can be performed in the usual way filtering by applying a filter window having a predefined size. However, in this implementation, the filtering is performed if the Central pixel of the filter window is the edge, and filtering is performed if the Central pixel of the filter window is not an edge. The filter box can be shared filter window. In this implementation is used chetyrehzvennoy filter box having five pixels arranged in the form of a cross with a single Central pixel, as shown in Fig.7. In Fig. 7 X represents the edge pixel, and a region other than regions with "X" are not edge pixels.

In addition, if the filter box does not have a boundary pixel, is conducted in a conventional filter, whereas if the edge pixel exists, proleva, ">" represents a shift to the right.

The invention may be embodied in digital General-purpose computer that executes the programs used in the computer media, including, but not limited to, such data carriers, magnetic data carriers (for example, ROM, floppy disks, hard disks, etc), optically readable media (e.g. CD-ROM, DVD, etc) and carrier waves (for example, transmission through Internet). Therefore, the present invention can be embodied as used in computer media embodied thereon computer-readable software code block for the signal adaptive filtering, and computer-readable software code means includes: computer-readable software code means to cause the computer to affect the production of the information block to reduce the effect of blocking and information edging for noise reduction outline of the coefficients of predetermined pixels of the upper and left edge regions of the data block when the frame obtained by decomposition of an image data stream of binary digits for inverse quantization, an inner frame; computer-readable the lock and information of the outline of the previous frame in accordance with the motion vector information block and the information of the outline of the current frame, if the frame is an intermediate frame, and to set the information edging to "1" that represents the request filtering for the image data, if there is a residual signal of the inverse quantized current block; and computer-readable software code means to cause the computer to act on adaptive filtering of the image data passing through based on the block inverse quantization and inverse discrete cosine transformation, in accordance with, for example, developed by the information block and the information of the border. Functional program, code and code segments used for the embodiment of the present invention can be obtained by a specialist in computer programming from the description of the invention contained herein.

Industrial applicability
As described above, the present invention can remove noise blocking and noise outline of the image restored from an image compressed on the basis of the blocks, thereby improving the image recovered from the compression.


Claims

1. The way the signal adaptive filtering can reduce the effect of blocking and shadowsa operations on which: (a) establish information horizontal block is equal to "1", which means the requirement of filtering for the image data, only when the ratio of the pixel And located in the upper left corner of the block is not equal to "0", or any coefficient of the pixels to the left boundary region of the block is not equal to "0" and set the vertical information of the block is equal to "1", which means the requirement of filtering for the image data, when only the coefficient of the pixel is not equal to "0" or any coefficient of the pixels of the upper boundary region of the block is not equal to "0", and establish information edging equal to "1", which means the requirement of filtering for the image data, when any coefficient of the pixels that are different from the pixels And to the right of the pixel A, and below the pixel A, is not equal to "0" when the frame obtained by decomposition of an image data stream of binary digits for inverse quantization, an internal frame, and (b) provide adaptive filtering image data according to the generated information block and the information of the border.

2. The method according to p. 1, in which the operation (a) further contains the installation information, blokirovki information outline of the current frame, if the frame is an inner frame, and installation information edging to "1", which represents the requirement of filtering for the image data, if there is a residual signal of the inverse quantized current block.

3. The method according to p. 2, in which in order to reduce the effect of blocking in the operation (b) horizontal filtering is performed using a weighted filter having a predetermined weighted value when the information of the horizontal block unit is "1" and information edging "0", and when the information of the horizontal block unit is "1" or information edging is not equal to "0", compare the absolute value of the difference between adjacent pixels and the value of Q used as the dividend for the quantization unit, and then filter with a preset value in accordance with the comparison result.

4. The method according to p. 2, in which in order to reduce the effect of blocking in the operation (b) vertical filtering is performed using a weighted filter having a predetermined weighted value when the information of the vertical block of the block is equal to "1" and information edging "0", and when the information of the vertical block of the block is equal to "1" or the s Q, used as the dividend for the quantization unit, and then filter with a preset value in accordance with the comparison result.

5. The method according to p. 3, in which the assumption that the block has a size 88 pixels, and pixels located at the center of the border block in window 16 are P1, P2, P3, P4, P5 and P6, to reduce the effects of blocking filtering performed on the pixels P1, P2, P3, P4, P5 and P6 on the left vertical boundary of the unit using a weighted filter having a weighted value (1,1,1,2,1,1,1), when the information of the horizontal block unit is "1" and information edging "0", and when the information of the horizontal block block is not equal to "1" or information edging is not equal to "0", the filtering is performed for the pixels P2, P3, P4 and P5 in accordance with the following algorithm
d=p4-p3;
if (ABS(d)Q){
P4=P4-(d/2); P3=P3+(d/2);
d=P5-P4;
if (ABS(d)Q) p5=p5-(d/4);
d=P3-P2;
if (ABS(d)Q) P2=P2+(d/4);
} else {
if (ABS(d/2)2Q)
if (d>0) {
p4=p4-(Q-ABS(d/2));
p3=p3+(Q-ABS(d/2));
}
else {
p4=p4+(Q-ABS(d/2));
p3=p3+(Q-ABS(d/2));}
d=p5-p4;
if (ABS(d)

6. The method according to p. 4, in which, assuming that the block size is 88 pixels, and pixels located at the center of the border unit in the window 61, are P1, P2, P3, P4, P5 and P6, to reduce the effects of blocking filtering performed on the pixels P1, P2, P3, P4, P5 and P6 on the upper horizontal boundary of the unit using a weighted filter having a weighted value (1,1,1,2,1,1,1), when the information of the vertical block of the block is equal to "1" and information edging "0", and when the information of the vertical locking block is not equal to "1" or information edging is not equal to "0", the filtering is performed for the pixels P2, P3, P4 and P5 in accordance with the following algorithm:
d=p4-p3;
if (ABS (d)Q) {
P4=P4-(d/2); P3=P3+(d/2);
d=P5-P4;
if (ABS(d)Q) P5=P5-(d/4);
d=p3-p2;
if (ABS(d)Q) p2=P2+(d/4);
} else{
if (ABS(d/2)<2Q)
p4=p4-(Q-ABS(d/2));
p3=p3+(Q-ABS(d/2));
}
else {
p4=p4+(Q-ABS(d/2));
p3=p3+(Q-ABS(d/2));
}
d=p5-p4;
if (ABS(d)Q) p5=p5-(d/4);
d is between blocks, ABS represents the absolute value, and "=" means that the value of the right side of the equation is replaced by the value of the left side.

7. The method according to PP.3, 4, 5 or 6 in which to reduce noise, fringing the method further includes the operation of checking whether information edging equal to "1", and performing a predetermined filtering, if the trim is set to "1", and passes the filter, if the trim is not set to "1".

8. The method according to p. 7, in which the predefined filter contains the following operations in which (x) perform the operation of calculating the gradient for the block, past the inverse quantization and inverse discrete cosine transformation using a predetermined one-dimensional horizontal and vertical gradient operators to define a boundary pixel, (y) to produce a binary edge map representing whether each pixel is an edge pixel, by using the absolute value of the difference between treated gradient operator value of one pixel and the value of the adjacent pixel, and the values of Q used as the dividend for the quantization unit, and (z) filter by applying a predetermined filter OK is ugenia vertical edges and operation (y2) detecting horizontal edges, and operation (y1) contains the following sub-operations, which provide: (y11) calculating the absolute value of A1 treated gradient operator values of the pixel P1, which is determining whether it is an edge pixel, and the pixel to the right of the pixel P1, (y12) calculating the absolute value of A2 processed gradient operator values of the pixel P1 and the pixel to the left of the pixel P1, (y13) comparison of absolute values of A1 and A2 with a preset threshold value Th, and determining whether the pixel P1 pixel vertical edges, in accordance with a Boolean value as the result of the comparison, and (y14) the implementation of the above sub-operations (y11)-(y13) for all pixels of the block, and the operation (y2) contains the following sub-operations, which provide: (y21) calculating the absolute value And'1 the difference between the treated gradient operator values of the pixel P1, which is determining whether it is an edge pixel block, and pixel below the pixel P1; (y22) calculation of absolute values And'2 the difference between the treated gradient operator values of the pixel P1 and the pixel above the pixel P1; (y23) comparison of absolute values of A'1 and a'2 with a preset threshold value T is receiving as a result of the comparison, and (y24) the implementation of the above sub-operations (y21)-(y23) for all pixels of the block.

10. The method according to p. 9, in which the sub-operation (y13) of the pixel P1 is determined as the edge pixel, and if true Boolean formula

and the pixel P1 is determined as not marginal, if this logical formula is false, but sub-operation (W) pixel P1 is determined as the edge pixel, and if true Boolean formula

and the pixel P1 is determined as not marginal, if this logical formula is false, where in the above logical formula "&&" represents a Logical And, and "" represents a Logical OR.

11. The method according to p. 8, in which in operation (z) after applying the filter window having a predefined size to the generated binary edge filtering is performed if the Central pixel of the filter window is an edge pixel, and the filtering is performed if the Central pixel of the filter window is not an edge pixel, and performs a normal filter, if the filter box does not have a boundary pixel, and performs a weighted filter, if the filter window has an edge pixel.

12. Method of filtering noise on the operation, on which (a) perform the operation of calculating the gradient on the unit is subjected to inverse quantization and inverse discrete cosine transform, using a predetermined one-dimensional horizontal and vertical gradient operators; (b) produce a binary edge map representing whether each pixel is an edge pixel by comparing the absolute value of the difference between treated gradient operator value of one pixel and the value of the adjacent pixel with a preset threshold value Th, and determine whether the pixel is an edge pixel, in accordance with the logical value of the comparison result; and (c) filter by applying a predetermined filter window to the generated binary edge map.

13. The method according to p. 12, in which the operation (y) contains the operation (y1) detection of vertical edges and operation (y2) detecting horizontal edges, and the operation (y1) contains the following sub-operations, which provide (y11) calculating the absolute value of A1 treated gradient operator values of the pixel P1, which is determining whether it is an edge pixel, and the pixel to the right of PI the flow to the left of the pixel P1, (y13) comparison of absolute values of A1 and A2 with a preset threshold value Th, and determining whether the pixel P1 pixel vertical edges, in accordance with a Boolean value as the result of the comparison, and (y14) the implementation of the above sub-operations (y11)-(y13) for all pixels of the block, and the operation (y2) contains the following sub-operations: (y21) calculating the absolute value And'1 the difference between the treated gradient operator values of the pixel P1, which is determining whether it is an edge pixel, and the pixel below the pixel P1, (y22) calculation of absolute values And'2 the difference between the treated gradient operator values of the pixel P1 and the pixel above the pixel P1, (y23) comparison of absolute values of A'1 and a'2 with a preset threshold value Th, and determining whether the pixel P1 pixel horizontal edges, in accordance with a Boolean value as the result of the comparison, and (y24) the implementation of the above sub-operations (y21)-(y23) for all pixels of the block.

14. The method according to p. 13, in which the sub-operation (y13) of the pixel P1 is determined as the edge pixel, and if true Boolean formula

and the pixel P1 is determined as not edge, e is logicheskie formula

and the pixel P1 is determined as not marginal, if the logical formula is false, where in the above logical formula "&&" represents a Logical And, and "" represents a Logical OR.

15. The method according to p. 12, in which in operation (z) after applying the filter window having a predefined size to the generated binary edge filtering is performed if the Central pixel of the filter window is an edge pixel, and the filtering is performed if the Central pixel of the filter window is not an edge pixel, and performs a normal filter, if the filter box does not have a boundary pixel, and performs a weighted filter, if filter box has edge pixel.

16. Signal adaptive filter, can reduce the effect of blocking and noise fringing in the image data when a frame is composed of blocks of a predetermined size containing the validation block mode flag to check the flag to determine whether the frame is an inner frame and an intermediate frame when the image data stream of binary digits to decompose the inverse quantization, the generator information about vnutri the filtering for the image data, when only the coefficient of the pixel And located in the upper left corner of the block is not equal to "0", or any coefficient of the pixels to the left boundary region of the block is not equal to "0", and setting information of the vertical block is equal to "1", which means the requirement of filtering for the image data, when only the coefficient of the pixel is not equal to "0" or any coefficient of the pixels of the upper boundary region of the block is not equal to "0", and setting information of the outline is "1", which means the requirement of filtering for the image data, when any coefficient of the pixels, different from the pixels And to the right of the pixel A, and below the pixel A, is not equal to "0" when the frame is defined by a block check mode flag as the inner frame; a generator of information about the interframe filter for installation information block and the information of the outline of the previous frame in accordance with the motion vector in the quality of the information block and the information of the outline of the current frame, if the frame is an intermediate frame, and installation information edging equal to "1" if there is a residual signal of the inverse quantized current block; and an adaptive filter for adaptively filtering the image data, the use of weighted filter, having a predetermined weighted value when the information block of the block is equal to "1" and information edging is equal to "0".

17. Signal adaptive filter according to p. 16, in which the adaptive filter performs a weighted filtering or filtering or not filtering in accordance with the information block and the information of the border.

 

Same patents:

The invention relates to techniques for television, in particular to the means of decoding the image signal

The invention relates to television technology, in particular to the transmission of video signals over a narrow-band channels, and for encoding broadband signals to reduce their bandwidth when the loss of information does not distort the overall perception

The invention relates to digital signal processing and can be used in the processing of radar images, in particular when processing two-dimensional response fields of radar, sonar, and optoelectronic sensors

The invention relates to the field of computer engineering and can be used for image processing

The invention relates to the field of computer engineering and can be used for image processing

The invention relates to techniques for television, in particular to the means of decoding the image signal

FIELD: technology for processing remote probing data for detection and recognition of objects on basis of their images.

SUBSTANCE: method includes source geometrically pixel-wise combined digitized images of one and the same scene concurrently in n spectrum ranges, source signs matrix is formed, each element of which represents n-dimensional values vector of signals of pixels of source images with similar coordinates, standard is selected in form of arbitrary element of source signs matrix, final numeric matrix is formed, to each current element of which value is assigned, equal to distance in vector signs space between vector, appropriate for standard and vector, appropriate for element of source matrix with same number of row and column to those of current element, final matrix is transformed to digital image, while as signs, brightness values of pixels are used, textural and gradient characteristics of source image pixels.

EFFECT: simplified operations concerning generation of synthesized image for visual interpretation, its adaptation to objects targeted by observer, detailed reflection of chosen objects on synthesized image and compact representation of information.

4 cl, 2 dwg

FIELD: technology for analysis of digital images by means of computing meshes.

SUBSTANCE: result is achieved by means of optimization and correction of grid nodes to provide extreme of compound function of spatial coordinates of grid nodes. Grid nodes are interpreted as atoms. Each node of grid provides a potential function for potential field of atom. Image represents a potential field. Compound function is a weighted total of potential fields of atoms and image, estimated in nodes of grid.

EFFECT: provision of possibility of grid initialization process and process of optimization of configuration of this grid relatively to characteristic features of image.

6 cl, 23 dwg

FIELD: processes for identifying pieces of art for the purpose of protecting from counterfeit and for determining extent of safekeeping.

SUBSTANCE: method involves analyzing palette of piece of painting on the basis of resulting sequence of digital data and/or comparing resulting sequence of digital data with data corresponding to original piece of art.

EFFECT: provision for analyzing of each individual color-divided image whose data are shown in the form of graphical image presenting in details gradation of each color shade on a portion of piece of art under investigation.

5 cl, 4 dwg

FIELD: digital processing of images, possible use for improving digital colored or halftone image.

SUBSTANCE: in accordance to invention, filtration of noises is performed, correction of global contrast is realized, brightness component is extracted from image as maximal one out of R, G, B components, local correction parameters are determined for light and dark tones, bilateral filtration of image brightness is performed, dark and light tones are corrected in brightness channel. Correction of dark image tones is performed due to addition to brightness channel of result of multiplication of difference between image of details in dark tones and brightness, inversion of result of bilateral filtration, raised to certain power, which determines width of tone spectrum, and amplification coefficient of dark tones, while image of details in dark tones is a function of ratio of brightness of original image and filtered image. Correction of light image tones is performed due to subtraction from brightness channel of difference between result of multiplication of brightness and image of details in light tones, result of bilateral filtration of brightness, raised to certain power which determines width of tone spectrum, and light tones weakening coefficient, while image of details in light tones is a function of ratio of filtered image and brightness of original image, image correction result is converted to RGB color system.

EFFECT: ensured correction of global contrast, local correction of light and dark tones, suppression of noises with a certain increase of image detail.

9 cl, 16 dwg

FIELD: technology for forming two-dimensional color images, for example, images (copies, reproductions) of color original, can be used for correction of color structure characteristics of images.

SUBSTANCE: in accordance to invention, copy of original is formed by applying colorant to copy carrier in accordance to interconnected set of planar and color coordinates of copy, while as planar coordinates of copy, planar coordinates of original are used, and color coordinates of copy are formed on basis of appropriate color coordinates of original by means of introduction of color-correcting corrections, determined during testing of copy of standard image produced on carrier.

EFFECT: increased quality of produced copies of original.

FIELD: generation of colored images.

SUBSTANCE: in accordance to the invention, data of image is stored, which image is supposed to be generated by each image generation block, in predetermined image data storage means; from predetermined means for storing information about exposition change value, showing value of shift relatively to scanning direction on image carrier of each image generator block, information is read about value of exposition change; coordinates of image data storage means reading address are transformed on basis of read information about exposition change value, and image data is read in accordance to information about transformed address; shade is corrected for pixel data, read at coordinate transformation stage based on information about transformed address; predetermined semitone processing is applied to image element data produced at correction stage; image element data, received during semitone processing stage, is outputted as signal for controlling exposition of exposition block of corresponding image generation block.

EFFECT: increased precision of reproduction.

6 cl, 35 dwg

FIELD: engineering of systems for analyzing digital images, and, in particular, systems for showing hidden objects on digital images.

SUBSTANCE: in accordance to the invention, method is claimed for visual display of first object, hidden by second object, where first object has color contrasting with color of second object, and second object is made of material letting passage of visible light through it, where amount of visible light passing through second object is insufficient for first object to be visible to human eye. The method includes production of digital image of first and second objects with usage of visible light sensor. Digital data of image, received by computer system, contains both data of first object and data of second object, where data of first object and data of second object contains color information, and value of contrast between first and second objects must amount to approximately 10% of full scale in such a way, that along color scale of 256 levels the difference equals approximately 25 levels, then data of second object is filtered, after that values, associated with data of first object, are increased until these values become discernible during reproduction on a display.

EFFECT: creation of the method for showing hidden objects in digital image without affecting it with special signals.

3 cl, 6 dwg

FIELD: engineering of systems for analyzing television images, in particular, for stabilizing an image in television images.

SUBSTANCE: in accordance to the invention, first digital image and at least second image have a set of pixels, and each pixel has associated address for display and is represented by color. System user sets a color matching interval, or system uses a predetermined color matching interval, then in first digital image a pixel is selected, for example, representing an element in an image, which is either fuzzy because of element movement, or appears trembling due to camera movement, and is matched within limits of interval with a pixel of second image. The interval ensures compensation, required during change of lighting. After selection of a pixel in first image, it may be matched with all pixels in the second image, where each pixel of the second image, having matching color within limits of matching interval, is stored in memory, and pixel color is selected, closest to pixel of first image. Then pixel addresses are changed in second image so that the address of pixel positioned in second image, closest color-wise to the pixel in the first image, is assigned the same address on the display as the pixel of first image and the resulting rearranged second image is dispatched into memory for storage.

EFFECT: creation of efficient image stabilization method.

9 cl, 11 dwg

FIELD: processing of digital images.

SUBSTANCE: method includes following operations: recording of digital images from, for example, a digital camera, into memory block, which may be a part of a computer or a printer; analysis of digital images, contained in memory block, on basis of properties: image brightness, image blur level, red eye detection; recording of metadata into files of digital images, corresponding to character of found defects; digital images, on basis of analysis results, are classified as belonging to following categories: images which satisfy high quality standards and which are usable for printing, in other words, without defects, images, usable for printing after being edited, in other words, with correctable defects, images, not usable for printing, in other words, with uncorrectable defects; in memory block, separate sections are created for storage of each one of three aforementioned categories of digital images; files of sorted digital images are recorded in memory block together with metadata about the character of found defects.

EFFECT: conservation of processing time and consumables.

6 dwg

FIELD: methods for removing noise in an image, possible use for improving quality of image.

SUBSTANCE: in accordance to the invention, effect is achieved due to conversion of brightness of image pixels with noise by means of solving the diffusion equation in non-divergent form, which ensures simultaneous suppression of noise and preservation of image edges.

EFFECT: simplified noise removal and increased quality of resulting digital image.

4 cl, 1 dwg

FIELD: video image processing.

SUBSTANCE: unit for pixel depth increasing is implemented with the possibility to scale colour pixel depth of video image frame, an area subdividing unit is implemented with the possibility to divide a video image frame producing areas of equal size for each colour and ensuring successive switching-over between areas and colours; at that, each conveyor is implemented with the possibility to delete false contours in two areas in four directions simultaneously, and an averaging unit is implemented with the possibility to calculate an average sum of data from outputs of four conveyors for each colour.

EFFECT: increase of image colour pixel depth along with deletion of false contours in real time.

16 cl, 3 dwg

FIELD: physics.

SUBSTANCE: invention is related to system and method for generating of image with expanded dynamic range (EDR) from grouped sequence of scene images, even with motion of scene or camera between images. Technical result consists in creation of better method for fixation of full radiation of static scene to create of EDR map. The said result is achieved due to the fact that one of images is selected as a reference image, every non-reference image is matched with one of the other images, as well as with reference image, which shows exposition, which is closer to this reference image than the one of considered image, and is closest among other images to exposition of considered image, generating field of flux. Fields of flux, which are generated for non-reference images, that are not any longer matched with reference image, are fixed for matching each of them with reference image, and every non-reference image later is changed with the help of flux field connected with it. Reference image and altered images are combined to create of luminosity chart that represents EDR image.

EFFECT: creation of better method for fixation of full radiation of static scene to create expanded dynamic range chart.

20 cl, 15 dwg

Up!