JPH01183276A - Adaptive filter system for picture coding transmission - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION [Summary] The present invention relates to an adaptive filter method for image encoding and transmission in which an image is encoded in block 11 using a required prediction method including intra-frame prediction and then transmitted.

The aim is to reduce block distortion that occurs at the edges of intra-frame predicted blocks in reconstructed images.

The receiving side is configured to perform filter processing such as interpolation on the pixels of the block for which intra-frame prediction has been performed.

[Industrial application field]

The present invention relates to an adaptive filter method for removing block distortion of a reproduced image in an image encoding transmission method in which an image is encoded block by block using a prediction method including at least intraframe prediction and then transmitted.

[Conventional technology]

As a video band compression encoding method for realizing video conferencing systems, etc., there is a known method in which images are encoded block by block. There is a method of encoding using any one of the prediction methods, or a method of adaptively selecting and encoding interframe prediction, motion compensation prediction, and intraframe prediction for each block according to the motion of the screen. Furthermore, the name of one invention related to the present applicant [September 2, 1986, called Variable Block Size Motion Compensation Method 1]
There is a method of encoding by adaptively selecting a prediction method and a block size according to the properties of each part of the five images, such as a still part or a moving part, as proposed in a patent application filed on the 2nd.

[Problem that the invention seeks to solve]

In these methods of encoding images in block units, when intra-frame prediction is used for block encoding, in the screen played back on the receiving side, the block for which intra-frame prediction Al+ has been performed is There is a tendency for block distortion to become noticeable, which makes it difficult to blend in with adjacent pros. Such block distortion tends to be noticeable even though the predicted error is not large, and it deteriorates the quality of the actual reproduced screen more than expected from the actual prediction error.

Therefore, the object of the present invention is to use an image encoding and transmission method (in which one image is encoded and transmitted into blocks l'n 417 using a required prediction method including intra-frame r-measurement). The objective is to reduce block distortion that occurs at the edges of the

[Failure to solve the problem]

FIG. 1 is a principle explanatory diagram for explaining an adaptive filter method for image coding transmission according to the present invention.

The adaptive filter method according to the present invention is an image coding transmission method in which nine images are encoded block by block using a required prediction method including intra-frame prediction, and transmitted. It is configured to perform filter processing on pixels.

The filter process uses the predicted value of a block subjected to intra-frame prediction and l'! of the block adjacent to that block. It is possible to use interpolation processing using the values.

It is also possible to perform interpolation only when the difference between the predicted value of a block that has been subjected to frame J/inner side as filter processing and the predicted value of the block adjacent to that block exceeds a threshold value. It is.

[Effect]

In the adaptive filter method according to the present invention, filter processing is performed on the receiving side only for block regions in which intra-frame prediction has been performed. (+) and (2) in Figure 1
, block 3 is a block subjected to intra-frame prediction, and block l.

It is assumed that 2 and 4 are blocks subjected to intra-frame prediction or other prediction.

As shown in FIG. 1 (1), when the value of a pixel at the edge of block 3 takes a value that is noticeably different from the value of a pixel at the edge of other adjacent blocks 1 and 2. 1 Block distortion will occur at this edge.

Therefore, in the present invention, the block 3 subjected to intra-frame prediction is subjected to filter processing as described below, for example, to suppress the occurrence of block distortion.

That is, when the difference between pixel ■ in block 3 and pixel ■ in block 4 exceeds a certain threshold, the value of pixel ■ is changed to block I'
1 between pixel ■ in Jtsuk3 and pixel ■ in Bmltsuk4
Interpolation processing is performed to make i average E. Also, if the difference between pixel ■ and pixel ■ in block l exceeds a certain threshold, pixel ■
Let the average value of pixel (2) and pixel (2) in block 3 be the value of pixel (2). Further, when the difference between pixel ■ and pixel ■ in block 2 exceeds a certain threshold value, the average value of pixel ■ and pixel ■ in block l is taken as the value of pixel ■.

Furthermore, as a result of the above-described processing, processing when distortion becomes noticeable even within block 3 will be described with reference to FIG. 1(2). In other words, within block 3, pixel ■
If the difference between and pixel ■ exceeds the threshold, pixel 1 and pixel 6
Let the average value of Φ be the value of pixel ■.

As described above, interpolation is performed only on block regions for which intra-frame prediction has been performed using the pixel values of that block and the pixel values of adjacent surrounding blocks.

〔Example〕

The present invention will be described in detail below with reference to one drawing.

FIG. 2 is a flowchart showing an adaptive filter method for image coding transmission as an embodiment of the present invention. In this embodiment, the screen is encoded and transmitted by adaptively selecting the required prediction method including intra-frame prediction in the middle of the block, and the size of the block to be encoded is as shown in Fig. 3. As shown, laterally Bx,! 1? BY direction, and the pixels within the block are denoted by F (i,j).

First, variables i+J indicating the pixel position are both set to 1 in step S2. S3), pixel of interest F(i, j)
It is determined whether the area belongs to the area in which intra-frame prediction has been performed (step 34). If it is outside the area, step S
Proceed to l+.

The calculation j=j+1 is performed, and if j is within the block size By (step 512), the process of step S4 is repeated.

If the pixel F (i, j) is within the area of intra-frame r measurement, Fl Set. This can be determined by whether or not the following conditional expression is satisfied (step S5).

1+1≦B× and i-1≧l This conditional expression applies to the left fp of the target pixel F(i, j)
By determining that the pixel position (i-1) of l+ is 1 or more and that the right pixel position (i41) is within the block size I'3x, the object angle AF (i, j) is set horizontally. This is to determine whether it is within the block screen in the axial direction.

If the target pixel F(i, j) is within the block screen,
Target pixel F (i, j) and its right pixel F (i, j)
or the absolute value of the difference between the target pixel F (i, j) and the pixel F (i-1, 3) to its left is greater than or equal to a predetermined threshold value of 78 using the following formula. (Step S6).

l F(i,j) −F(i+i,j) l
≧T111 F(i,j) −F(i−1,j
) l ≧Ttl If these differences exceed the threshold (/jTl+, interpolation processing in the horizontal axis direction is performed for the target pixel F (i, j) in step S7), and from the L threshold TH If it is smaller, proceed to step 311.

The interpolation process is performed by finding the average value of the pixels on the left and right of the target pixel F (i, j) using the following equation and setting it as the value of the target pixel F (i, j).

F(i,j) −'A (F(i+1.j) +F
(i-1, j)) After performing this interpolation process in the horizontal axis direction, it is further determined whether the target pixel F(i, j) protrudes upward or downward from both sides of the block in the vertical axis direction. judge. This can be determined by whether or not the following conditional expression is satisfied (step S8).

j41'; By and j-1≧1 This conditional expression indicates that the lower pixel position (j-1) of the target pixel F (i, j) is 1 or more, and - the right pixel position ( j+]) is within the block size BY, thereby determining that the target pixel F(i, j) is within the block screen in the vertical axis direction.

If the target pixel F (i, j) is within the professional screen,
Target pixel F(i,j) and its upper pixel F(i,j)I
) or the absolute value of the difference between target steel: JF (i, j) and its lower pixel) (i, j-1) is less than a predetermined threshold Tll? is determined by the following equation (step S9).

1F(i,j) F(i,j+1) l≧Tl1l
F(i,j) -F(i,j-1) l≧TH If these differences exceed the threshold Tll, the target pixel F
It is assumed that (i, j) is subjected to interpolation processing in the vertical axis direction (step 5IO), and if it is smaller than the threshold Tl+, the process proceeds to step Sll.

The interpolation process in the vertical axis direction is performed by finding the average value of the pixels L below the target pixel F(i, j) using the following equation and setting it as the value of the target pixel F(i, j).

F(i, j) ='A (F(i, j+1) + F
(i, j-1)) After performing this interpolation process in the vertical axis direction, the value of j is increased by one and the above operations are repeated (steps S4 to s+2). This process is performed when the value of j is By repeating this until the block size By is exceeded (step 5I2), the interpolation process is performed for all pixels in the vertical axis direction within the block screen at the position i in the horizontal axis direction.

When the vertical axis position of the pixel at the horizontal axis direction position i exceeds the size By (step 5I2).

Next, increase the value of i by one (step 313),
The value of j is reset to 1 (step S3), and the above operations are repeated (steps 83 to 5I4).

This process is repeated until the value of i (i.e., the position of the pixel in the horizontal axis direction) exceeds the block size Bx (step 31
4).

As a result, it is possible to perform interpolation processing in the vertical and horizontal directions for all pixels within the area of block size RxXBy, thereby reducing block distortion that occurs between adjacent blocks or within the own block. It is possible.

Various modifications are possible in carrying out the invention. For example, in the above embodiment, the case where interpolation is performed in the vertical axis direction and the horizontal axis direction is described as "ζ".

Of course, the invention is not limited to this, and it is also possible to perform the interpolation in an oblique direction. Further, in the above-described embodiment, the interpolation process is performed when the difference between the target pixel and the pixels located above, below, or to the left or right of the target pixel exceeds a predetermined threshold value, but the present invention is not limited to this.

Interpolation processing may be uniformly performed on all pixels without using the concept of threshold value.

〔Effect of the invention〕

According to the present invention, since filter processing is performed on the part where intra-frame prediction has been performed, the block distortion is not 1. can be regenerated.

[Brief explanation of the drawing]

FIG. 1 is a principle explanatory diagram for explaining an adaptive filter method for image coding transmission according to the present invention. FIG. 2 is a flowchart showing the processing procedure of an adaptive filter method for image coding and transmission as an embodiment of the present invention. and. FIG. 3 is a diagram illustrating blocks to be encoded. In fig. 1 to 4-Block■~[Phase] One pixel () Shim Uchiko;'all) Flock 1 Flock 2 Ll amount S Sun Moon IE denial 2 苧 - Katana Shi Tomoko - Akira Shia
Figure 1 ex Figure 3

Claims (1)

[Claims] 1. In an image encoding transmission method in which an image is encoded block by block using a required prediction method including intraframe prediction and transmitted, on the receiving side, the pixels of the block subjected to intraframe prediction are An adaptive filter method configured to perform filter processing based on 2. The adaptive filter method according to claim 1, wherein the filtering process is a process of performing interpolation using a predicted value of a block subjected to intra-frame prediction and a predicted value of a block adjacent to the block.