JPH1098722A - Block distortion removing filter, image processor, and filtering method for image signal - Google Patents

Abstract

PROBLEM TO BE SOLVED: To sufficiently remove even block distortion which is small in a difference value of a signal at a border of a block without spoiling a high--frequency component of a source image. SOLUTION: When a pixel signal included in pixel blocks forming one image is supplied, a detection part 1 detects a difference in pixel signal level at the periphery of the border of pixel blocks. This difference is compared by a threshold value decision part 2 with a reference value. According to the result of the comparison, an adder 5 adds a specific addition value to the pixel level at the periphery of the said block border. When the specific addition value is added, a random number is generated and based on this random number, a pattern containing the said specific addition value is selected out of addition value patterns. Consequently, the regularity of the difference value at the border of the pixel blocks is disordered and the block distortion generated between pixel blocks can be removed.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a filter for removing block distortion, an image processing apparatus, and a method for filtering an image signal. In particular, the present invention relates to a block distortion removal filter for removing block distortion generated at the time of, for example, compression encoding or decoding of an image, and an image processing apparatus using the same.

[0002]

2. Description of the Related Art When an image is compressed with high efficiency using an image compression coding technique, image quality deterioration peculiar to various compression coding techniques often occurs. For example, in the case of an encoding method in which image compression is performed in block units using DCT (Discrete Cosine Transform) or the like, block distortion may occur in which block boundaries are discontinuous.

[0003] Block distortion is the most visually noticeable deterioration in image quality in a compression coding technique that performs processing in units of blocks, and it is necessary to reduce or eliminate this in order to improve image quality. For this purpose, a conventionally performed method is to locally or globally apply a low-pass filter to a decoded image containing block distortion.

FIG. 8 is a diagram for explaining a process of removing block distortion by a conventional low-pass filter.
FIG. 8A shows the luminance value of the decoded image before the filter processing, and FIG. 8B shows the luminance value after the filter processing. In these figures, four scan lines orthogonal to the block boundaries in the image are shown as representatives, and the vertical direction of each line indicates the luminance level of the pixels arranged on each line indicated by a black circle. Is represented.

As shown in FIG. 8A, a difference Δ occurs in the luminance level uniformly in each line orthogonal to the block boundary indicated by the dotted line (the value of Δ may be slightly different for each line. , Almost the same value), and block distortion causing discontinuous block boundaries occurs. If there is such a block distortion, the decoded image looks like a mosaic at that location, and it becomes considerably hard to visually recognize.

Therefore, conventionally, as shown in FIG.
The difference Δ in the luminance level at the block boundary is dulled by the low-pass filter, and the block distortion is reduced by eliminating the linear edge of the luminance value.

[0007]

However, when the block distortion removal processing is performed using a low-pass filter,
There is a problem that high-frequency components of the original image other than the block distortion are also removed, and the image becomes blurred. In addition, when the value of the difference Δ at the block boundary shown in FIG. 8A is small, the effect of dulling the linear edge indicated by the difference Δ becomes small, and the block distortion cannot be removed effectively. There was also.

As an example of the latter problem, luminance 2
When the value of the difference Δ is only “1” in the 56-gradation image, the effect of the low-pass filter cannot be expected at all because the difference Δ has no intermediate value, and the block distortion is visually perceived clearly. Had been done. In this manner, even block distortion with a small difference Δ is often annoying, and therefore, in the related art in which block distortion with a small difference Δ cannot be removed, image quality degradation due to block distortion can be sufficiently suppressed. Did not.

SUMMARY OF THE INVENTION The present invention has been made to solve such a problem, and it is possible to remove block distortion without impairing the high-frequency component of the original image, and furthermore, to obtain a signal difference at a block boundary. An object of the present invention is to provide a block distortion removal filter and a removal method capable of sufficiently removing even a small block distortion.

[0010]

A block distortion removing filter according to the present invention is provided with a pixel signal included in a plurality of pixel blocks forming one image and receives at least two pixels around a block boundary of the plurality of pixel blocks. Detecting means for detecting a difference between the pixel signal levels of the pixel signal, comparing means for comparing the difference with a reference value, and adding means for adding a predetermined value to the pixel signal level around the block boundary according to a result of the comparison It is characterized by having.

According to another feature of the present invention, there are provided a generating means for generating a random number, and a selecting means for selecting a pattern including the predetermined value from a plurality of predetermined value patterns based on the random number. It is characterized by.

Here, the comparing means may compare the absolute value of the difference with the reference value. Further, the adding means, when the difference is smaller than the reference value,
The predetermined value may be added to the pixel signal level around the block boundary. Further, the random number may be a 2-bit pseudo random number.

An image processing apparatus according to the present invention is provided with decoding means for decoding pixel signals included in a plurality of pixel blocks forming one encoded image, and receiving the decoded pixel signals, Detecting means for detecting a difference between at least two pixel signal levels around a block boundary of the plurality of pixel blocks; comparing means for comparing the difference with a reference value;
Adding means for adding a predetermined value to the pixel signal level around the block boundary according to the result of the comparison.

According to another feature of the present invention, an encoding means for encoding a prediction error signal of a pixel signal included in a plurality of pixel blocks forming one image, and a decoding means for decoding the encoded pixel signal Decoding means for converting the pixel signals into a plurality of pixel blocks and detecting a difference between at least two pixel signal levels around a block boundary of the plurality of pixel blocks; Comparing means for comparing a pixel signal level around the block boundary with a predetermined value according to a result of the comparison, a predicting means for generating a prediction signal from an output signal of the adding means, And a difference unit that calculates a difference between a pixel signal included in a plurality of pixel blocks forming one image and the prediction signal and generates the prediction error signal.

Another feature of the present invention is that decoding means for decoding pixel signals included in a plurality of pixel blocks forming one encoded image, A first adding means for adding the signals and an output signal of the first adding means, and at least two signals around a block boundary of the plurality of pixel blocks are provided.
Detecting means for detecting the difference between the pixel signal levels, comparing means for comparing the difference with a reference value, and adding a predetermined value to the pixel signal level around the block boundary according to the result of the comparison. And an estimation means for generating the prediction signal from the output signal of the second addition means.

According to the image signal filtering method of the present invention, a pixel signal included in a plurality of pixel blocks forming one image is supplied, and a difference between at least two pixel signal levels around a block boundary of the plurality of pixel blocks is provided. And comparing the difference with a reference value, and adding a predetermined value to a pixel signal level around the block boundary according to a result of the comparison.

According to another feature of the present invention, the method further includes a step of generating a random number and selecting a pattern including the predetermined value from a plurality of predetermined value patterns based on the random number.

Here, the comparing step may include a step of comparing the absolute value of the difference with the reference value. Further, the adding step may include a step of adding the predetermined value to a pixel signal level around the block boundary when the difference is smaller than the reference value. Also,
The random number may be a 2-bit pseudo random number. The supplying step may include a step of supplying a pixel signal included in a plurality of pixel blocks forming the one image for each scan line. Further, the scan line may be a scan line in a horizontal direction or a vertical direction with respect to the one image.

According to another feature of the present invention, in the supplying step, the pixel signals included in the plurality of pixel blocks forming the one image are converted for each horizontal scan line of the one image and in the one image. Each is supplied for each vertical scan line, and a difference between at least two pixel signal levels around a block boundary of the plurality of pixel blocks in each of the horizontal and vertical directions is detected, and the difference and a reference value are detected. Comparing, and adding a predetermined value to a pixel signal level around the block boundary according to a result of the comparison.

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a diagram showing an embodiment of a block distortion removing filter according to the present invention. The block distortion removal filter shown in FIG.
, A threshold value determination unit 2, a pattern selection unit 3, a random number generation unit 4, and an adder 5.

The block boundary difference detecting section 1 generates a signal value (for example, a signal at a block boundary) in an image divided into a plurality of blocks (for example, a decoded image obtained by decoding an image encoded in units of blocks using DCT or the like). The difference between the luminance signal values is detected. Further, the threshold value determination unit 2 compares the absolute value of the detected difference with a predetermined threshold value, and determines whether or not to perform the block distortion removal processing at the corresponding block boundary.

The pattern selector 3 randomly selects a pattern of an added value to be added to signal values of a plurality of pixels around a block boundary on a scan line orthogonal to the block boundary from a plurality of patterns. Random number generator 4
Generates a 2-bit pseudo-random number, for example. The pattern selection unit 3 randomly selects a pattern from a plurality of patterns according to the generated random numbers. When the threshold value determination unit 2 determines that the block distortion removal processing is to be performed, the adder 5 adds the signal values of a plurality of pixels around the block boundary selected by the pattern selection unit 3 to each of the signal values. Add values.

The operation of the block distortion removal filter shown in FIG. 1 will be described below with reference to FIGS.
As shown in FIG. 2A, when a raster scan is performed orthogonally to a block boundary, it is assumed that there is a block boundary indicated by a dotted line between pixel positions n and n + 1 on a scan line. In this case, the difference value Δ of the example luminance signals of both pixels in the block boundary, the pixel value at a pixel position n X _n
Δ = _{Xn + 1− Xn} , which is obtained by the block boundary difference detection unit 1 in FIG. 1 in step S1 in FIG.

Next, in step S2, the absolute value | Δ | of the difference value Δ is compared with a predetermined threshold value Th by the threshold value determination section 2. When the absolute difference value | Δ | is smaller than the threshold value Th, it is assumed that block distortion exists, and block distortion removal processing after step S3 is performed. On the other hand, when the absolute difference value | Δ | is equal to or greater than the threshold value Th, the difference between the pixel values is regarded as an edge component existing in the original image, and the luminance value on the scan line is output as it is.

Since the block distortion is noise that is supposed to be absent, the difference absolute value | Δ | is often smaller than the edge component existing in the original image. Therefore, even if the block distortion removal processing is performed only when the absolute difference value | Δ | is smaller than the threshold value Th, and the processing is excluded when the absolute difference value | Δ | Can be removed.

Rather, if the processing is performed even when the difference absolute value | Δ | is large, the noise generated by randomization of the luminance value increases, which may itself lead to deterioration of the image quality. Therefore, it is possible to uniformly perform processing on all block boundaries. However, by performing a threshold determination as in the present embodiment, most block distortions can be removed, and block distortion removal can be performed. By performing the processing, it is possible to prevent the disadvantage that the image quality is rather deteriorated.

If it is determined in step S2 that the absolute difference value | Δ | is smaller than the threshold value Th and block distortion exists, the threshold value determination unit 2 determines in step S3.
In step (2), the pattern selection unit 3 determines the signal values of a plurality of pixels around the block boundary on one scan line (in the present embodiment, pixel position n−1) according to, for example, a 2-bit pseudo random number generated by the random number generation unit , N, n + 1, n +
One of the following four patterns is randomly selected from the following four patterns.

Pattern X _n−1 X _n X _{n + 1} X _{n + 2} 1) + Δ 0 −Δ 0 2) 0 + Δ 0 −Δ 3) 0 + Δ −Δ 0 4) + Δ 0 0 −Δ

Then, in step S4, the adder 5
The added value of the pattern selected by the pattern selection unit 3 is converted into four pixel values X _n−1 , X _n , X _{n + 1} , X around the block boundary.
_{Adds to n + 2} and outputs the result. For example, when the pattern 1) is selected, the pixel value X _n-1 + delta is, - [delta is outputted are respectively added to the pixel value X _{n + 1,} the pixel value X _n, X _{n + 2} Is output as is.

After the above-described pattern selection and addition processing is completed for one scan line, the same processing is performed for the next scan line. When such processing is performed for each scan line (four in the example of FIG. 2A), as shown in FIG. 2B, a block is selected by randomly selecting a pattern in each scan line. The linear edge of the signal difference value at the boundary disappears,
Block distortion is removed.

In this embodiment, the probability of continuously selecting the same pattern for two adjacent scan lines on the same block boundary is 1/4. As described above, since the probability of selecting the same pattern continuously is low in the block distortion removal filter of the present embodiment, the block distortion is effectively removed by disrupting the regularity of the signal difference value at the block boundary. Note that there is a possibility that the same pattern is continuously selected, but there is no visual problem because the probability is low.

In this embodiment, the difference value Δ is equal to the threshold value Th.
The above-described block distortion removal processing is performed only when the size is smaller than the above, and the processing target is limited to a narrow range of two pixels on both sides of the block boundary. It is possible to remove the block distortion without using. In an image having a luminance signal value as shown in FIG. 2B, the value is disturbed near the boundary of the block, but since the difference value Δ is small, there is no problem visually. I have. The processing target is
Not only two pixels on both sides of the block boundary but also three or more pixels may be used. The probability that three or more pixels will not select the same pattern continuously increases, and the block distortion removal performance increases.

Further, in this embodiment, since the block distortion is removed by disturbing the linearity of the edge along the boundary of the pixel block, when the difference value Δ is extremely small (particularly, when Δ = 1, ) Can also effectively remove the block distortion.

The block distortion removed in the processing of this embodiment is in the direction orthogonal to the scan line as shown in FIGS. Therefore, for example, in the case of a rectangular block, the processing for the horizontal scan line and the processing for the vertical scan line are performed in a cascade to cope with both horizontal and vertical block distortions. Can be.

Here, the removal of block distortion in the horizontal and vertical directions of an image signal will be described with reference to FIG. FIG. 4 is an example to which the embodiment of FIG. 1 is applied. FIG.
In the figure, reference numeral 6 denotes an image memory for storing an externally input image signal. A scanning circuit 7 scans the image signal from the image memory 6 in the horizontal and vertical directions and outputs a luminance signal of the image. Reference numeral 8 denotes a control circuit for instructing the scanning circuit 7 to scan in the horizontal or vertical direction. 9 is a block distortion removal filter,
For example, it is configured by the circuit shown in FIG.

Next, the operation will be described. The image signal stored in the image memory 6 is supplied to the scanning circuit 7.
The control circuit 8 instructs the scanning circuit 7 to switch the scanning direction of the input image signal to the horizontal direction. As a result, the scanning circuit 7 scans the input image signal in the horizontal direction, and outputs the luminance signal of the horizontal image to the block distortion removal filter 9. The block distortion removal filter 9 processes the input luminance signal and feeds it back to the image memory 6.

When the scanning in the horizontal direction is completed, the control circuit 8 controls the scanning circuit 7 to perform the scanning in the vertical direction.
Is instructed to switch the scanning direction of the input image signal to the vertical direction. Then, the scanning circuit 7 scans the input image signal in the vertical direction in the same manner as in the horizontal scanning,
The luminance signal of the image in the vertical direction is output to the block distortion removal filter 9. The block distortion removal filter 9 processes the input luminance signal and feeds it back to the image memory 6.

In the above example, the block distortion in the horizontal direction is removed first, and then the block distortion in the vertical direction is removed. However, the order may be reversed.

The application examples in which the above-described block distortion removal filter of the present embodiment is applied to moving picture coding without using still picture coding or prediction and when applied to moving picture predictive coding are described below. This will be described with reference to FIGS. FIG. 5 shows a configuration example of a device corresponding to moving image coding without using still image coding or prediction, and FIGS. 6 and 7 show a configuration example of a device corresponding to moving image prediction coding.

In the case of still picture coding or moving picture coding without using prediction, as shown in FIG. Apply as Image data (pixel signals included in a plurality of pixel blocks forming one image) encoded by an encoding device (not shown) is decoded by a decoder 10 and post-filter 11 is decoded.
(Block distortion removal filter) to remove block distortion occurring in the image. Here, when the encoding is performed in units of rectangular blocks, the block distortion removal filter of the present embodiment performs the processing on the horizontal scan lines and the processing on the vertical scan lines in a cascade.

In the case of video predictive coding, the decoder 10 and the post-filter 11 shown in FIG. 5 may be provided at the subsequent stage of the encoder not shown, but as shown in FIG.
The block distortion removal filter of this embodiment can be used as a loop filter 15 for the output of the local decoder 14. In this case, image data (pixel signals included in a plurality of pixel blocks forming one image), which are prediction errors encoded by the encoder 13, are decoded by the local decoder 14, and the decoded image is decoded. The data is,
After the block distortion is removed by a loop filter 15 (block distortion removal filter), the block distortion is supplied to the predictor 16. The prediction signal generated by the predictor 16 is
Given to. The differentiator 12 compares the input image signal with the prediction signal, and outputs the difference (prediction error) signal to the encoder 13.
Give to.

A decoding device corresponding to the encoding device shown in FIG. 6 is configured as shown in FIG. Image data (pixel signals included in a plurality of pixel blocks forming one image) as prediction errors encoded by the encoder 13 of FIG. 6 is decoded by the decoder 17 shown in FIG. The decoded image data and the prediction signal from the predictor 20 are added by the adder 18 and the block distortion removing filter 1 of the present embodiment is added.
The signal is fed back to the predictor 20 after the block distortion is removed.

As shown in FIGS. 6 and 7, by applying the block distortion removal filter of the present embodiment to both the encoder side and the decoder side, the decoder side as shown in FIG. It is possible to obtain a stronger block distortion removing effect as compared with a case where the filter is simply applied as a post filter.

[0044]

As described above, the filter for removing block distortion according to the present invention first detects a difference between pixel signal values around a block boundary of a plurality of pixel blocks forming one image.
This difference is compared with a reference value. Then, according to the result of the comparison, a predetermined addition value is added to the pixel signal values around the block boundary. The addition value is randomly selected from a plurality of predetermined addition value patterns. The smaller the probability that the same addition value pattern is selected in adjacent scan lines is, the more the regularity of the difference value at the boundary of the pixel block is disturbed, and the block distortion can be effectively removed. Therefore, block distortion with a small difference in pixel value at the block boundary can be effectively removed without damaging the high frequency components of the original image.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an embodiment of a block distortion removal filter according to the present invention.

FIG. 2 is a diagram for explaining the operation of the block distortion removal filter of the present embodiment.

FIG. 3 is a flowchart illustrating an operation of a block distortion removal filter according to the embodiment.

FIG. 4 is a diagram illustrating an example in which the block distortion removal filter according to the present embodiment is applied to image signal processing.

FIG. 5 is a diagram illustrating an application example of the block distortion removal filter of the present embodiment.

FIG. 6 is a diagram illustrating an application example of the block distortion removal filter of the present embodiment.

FIG. 7 is a diagram illustrating an application example of the block distortion removal filter of the present embodiment.

FIG. 8 is a diagram for explaining an operation of a conventional block distortion removal filter.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Block boundary difference detection part 2 Threshold judgment part 3 Pattern selection part 4 Random number generation part 5 Adder 6 Image memory 7 Scanning circuit 8 Control circuit 9 Block distortion removal filter 10 Decoder 11 Post filter (Block distortion removal filter) DESCRIPTION OF SYMBOLS 12 Difference machine 13 Encoder 14 Local decoder 15 Loop filter (block distortion removal filter) 16 Predictor 17 Decoder 18 Adder 19 Block distortion removal filter 20 Predictor

Claims (17)

[Claims] Detecting means for receiving a pixel signal included in a plurality of pixel blocks forming one image and detecting a difference between at least two pixel signal levels around a block boundary of the plurality of pixel blocks; A block distortion removing filter, comprising: comparing means for comparing the difference with a reference value; and adding means for adding a predetermined value to a pixel signal level around the block boundary according to a result of the comparison. 2. The apparatus according to claim 1, further comprising: generating means for generating a random number; and selecting means for selecting a pattern including the predetermined value from a plurality of predetermined value patterns based on the random number. Block distortion removal filter. 3. The filter according to claim 1, wherein the comparing unit compares the absolute value of the difference with the reference value. 4. The block distortion removal filter according to claim 1, wherein the adding means adds the predetermined value to a pixel signal level around the block boundary when the difference is smaller than the reference value. . 5. The filter according to claim 2, wherein the random number is a 2-bit pseudo random number. 6. A decoding means for decoding a pixel signal included in a plurality of pixel blocks forming one encoded image, and a decoding means for receiving the decoded pixel signal, Detecting means for detecting a difference between at least two pixel signal levels around the block boundary; comparing means for comparing the difference with a reference value; and determining the pixel signal level around the block boundary according to a result of the comparison. An image processing apparatus comprising: an adding unit that adds a predetermined value. 7. Encoding means for encoding a prediction error signal of a pixel signal included in a plurality of pixel blocks forming one image, decoding means for decoding the encoded pixel signal, and decoding A detecting means for receiving a converted pixel signal and detecting a difference between at least two pixel signal levels around a block boundary of the plurality of pixel blocks; a comparing means for comparing the difference with a reference value; Adding means for adding a predetermined value to the pixel signal level around the block boundary according to the result of the comparison; predicting means for generating a prediction signal from an output signal of the adding means; and a plurality of pixel blocks forming the one image An image processing apparatus comprising: a difference unit that obtains a difference between a pixel signal included in the prediction signal and the prediction signal and generates the prediction error signal. 8. A decoding means for decoding a pixel signal included in a plurality of pixel blocks forming one encoded image, and a first means for adding the decoded pixel signal and a prediction signal.
Adding means provided with the output signal of the first adding means, detecting means for detecting a difference between at least two pixel signal levels around block boundaries of the plurality of pixel blocks; A second adding means for adding a predetermined value to a pixel signal level around the block boundary in accordance with a result of the comparison, and a prediction signal from the output signal of the second adding means. An image processing apparatus comprising: a prediction unit configured to generate an image. 9. A method for supplying a pixel signal included in a plurality of pixel blocks forming one image, detecting a difference between at least two pixel signal levels around a block boundary of the plurality of pixel blocks, Comparing the value with a predetermined value to a pixel signal level around the block boundary according to a result of the comparison. 10. The filtering of an image signal according to claim 9, further comprising a step of generating a random number and selecting a pattern including the predetermined value from a plurality of predetermined value patterns based on the random number. Method. 11. The method according to claim 9, wherein the comparing step includes a step of comparing the absolute value of the difference with the reference value. 12. The image according to claim 9, wherein the adding step includes a step of adding the predetermined value to a pixel signal level around the block boundary when the difference is smaller than the reference value. How to filter the signal. 13. The method according to claim 10, wherein the random number is a 2-bit pseudo random number. 14. The image signal filtering method according to claim 9, wherein said supplying step includes a step of supplying pixel signals included in a plurality of pixel blocks forming said one image for each scan line. Method. 15. The method according to claim 14, wherein the scan line is a scan line in a horizontal direction with respect to the one image. 16. The method according to claim 14, wherein the scan line is a scan line in a direction perpendicular to the one image. 17. The supplying step supplies a pixel signal included in a plurality of pixel blocks forming the one image to each horizontal scan line of the one image and each vertical scan line of the one image. Detecting a difference between at least two pixel signal levels around a block boundary of the plurality of pixel blocks in each of the horizontal and vertical directions; comparing the difference with a reference value; 10. The method according to claim 9, further comprising the step of adding a predetermined value to the pixel signal level around the block boundary.