CN100345450C - Deblocking filtering method and apparatus of video frequency or image - Google Patents

Deblocking filtering method and apparatus of video frequency or image Download PDF

Info

Publication number
CN100345450C
CN100345450C CNB2005100490020A CN200510049002A CN100345450C CN 100345450 C CN100345450 C CN 100345450C CN B2005100490020 A CNB2005100490020 A CN B2005100490020A CN 200510049002 A CN200510049002 A CN 200510049002A CN 100345450 C CN100345450 C CN 100345450C
Authority
CN
China
Prior art keywords
filtering
block
macro block
elimination filtering
macro
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CNB2005100490020A
Other languages
Chinese (zh)
Other versions
CN1652610A (en
Inventor
虞露
易峰
董洁
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
XFusion Digital Technologies Co Ltd
Original Assignee
Zhejiang University ZJU
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Zhejiang University ZJU filed Critical Zhejiang University ZJU
Priority to CNB2005100490020A priority Critical patent/CN100345450C/en
Publication of CN1652610A publication Critical patent/CN1652610A/en
Application granted granted Critical
Publication of CN100345450C publication Critical patent/CN100345450C/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Abstract

The present invention discloses a deblocking filtering method for a video or an image, and a device thereof. The deblocking filtering method adopted by the present invention utilizes only the type and the quantization parameters of macro blocks for determining whether a deblocking filtering operation is carried out, and selecting an adopted filtering mode of the macro blocks when the deblocking filtering operation is needed, and the deblocking filtering method does not utilize the residual information or the motion vector information of the coded macro blocks for determining whether the deblocking filtering operation is carried out, or selecting the adopted filtering mode of the macro blocks when the deblocking filtering operation is needed; in addition, the procedure of the deblocking filtering operation is determined by automatic adaptation according to the filtering mode of the macro blocks and the content of the image. By the deblocking filtering method adopted by the present invention, the complexity of calculation and the complexity of storage is reduced on the premise that the objective quality and the subjective quality of the decoded image are guaranteed to be equivalent to that of the existing deblocking filtering method; the present invention is not limited in the technical field of the coding or the decoding of the video or the image, and can be popularized to other applications for signal processing.

Description

The method and apparatus of video or image block elimination filtering
Technical field
The present invention relates to a kind of electric numerical data treatment technology, especially, relate to a kind of method and apparatus that applies to video or image block elimination filtering.
Background technology
H.261 traditional video encoding standard such as the ITU of International Telecommunications Union formulate, H.263, H.26L, H.264 the MEPG-1 that MPEG formulates knits in the Motion Picture Experts Group of standard and the ISO of International Standards Organization, MPEG-2, MPEG-4 etc. are based on the hybrid coding Hybrid Coding framework.So-called hybrid encoding frame is to take all factors into consideration prediction, and the coding framework of the method for conversion and entropy coding has following main feature: 1, utilize prediction to remove the redundancy of time-domain; 2, utilize conversion to remove the redundancy of spatial domain; 3, utilize entropy coding to remove statistical redundancy.
In existing hybrid encoding frame, prediction and conversion are based on all that block of pixels carries out.When excessive or forecasting process result was not accurate enough when the quantization degree of conversion coefficient, blocking artifact can appear in decoded picture, caused subjective quality to descend.Can eliminate blocking artifact effectively by decoded picture being carried out block elimination filtering, improve subjective quality.H.264, the block-removal filtering method that up-to-date in the world video encoding standard adopts, computation complexity be than higher, account for usually a decoder computation complexity 1/3rd in addition more than.This mainly is because in order to eliminate blocking artifact to greatest extent when being avoided level and smooth real border, block-removal filtering method need utilize the coded macroblocks residual error, the information of each side such as motion vector and macro block (mb) type is determined the block elimination filtering process according to the picture material self adaptation, and amount of calculation is very big.Simultaneously, in the block elimination filtering process, need on piece one-level and pixel one-level, carry out repeatedly condition judgment, therefore be difficult to utilize the parallel computation function that provides as hardware platforms such as DSP.Secondly, because the number of the sampled pixel point that the block elimination filtering process relates to is a lot, all pixels of a two field picture all will pass through block elimination filtering basically, so the read-write control of internal memory etc. is operated also very consuming time and complicated.
Summary of the invention
The purpose of this invention is to provide a kind of method and apparatus that applies to video or image block elimination filtering.The block-removal filtering method that the present invention adopts has under the prerequisite of suitable objective and subjective decoded image quality in assurance and existing block-removal filtering method, greatly reduces computation complexity and storage complexity.
In order to achieve the above object, the technical solution used in the present invention is as follows:
The method of a kind of video or image block elimination filtering, (1) only utilize the quantization parameter of macro block (mb) type and macro block to determine whether to carry out block elimination filtering and the macro block filtering pattern of selection employing when needing block elimination filtering, and the macro block filtering pattern of not utilizing coded macroblocks residual information or motion vector information to determine whether to carry out block elimination filtering and when needs carry out block elimination filtering, selecting to adopt; (2) determine the block elimination filtering process according to macro block filtering pattern and picture material self adaptation.
The device of a kind of video or image block elimination filtering, it comprises: a macro block filtering mode adjudging device, according to macro block (mb) type, quantization parameter, the quantization parameter threshold value is adjudicated current filter patterns, the filter patterns of decision current macro; One deblocking filtering apparatus carries out block elimination filtering according to the output of filter patterns decision device to the current macro coded data and handles; One time-division multiplexer; Wherein, the filter patterns decision device links to each other with deblocking filtering apparatus, and deblocking filtering apparatus links to each other with TDMD time division multiplex device.
The present invention compares with background technology, the beneficial effect that has: the macro block filtering pattern that quantization parameter determines whether to carry out block elimination filtering and selection is adopted when needing block elimination filtering of only utilizing macro block (mb) type and macro block, and the macro block filtering pattern of not utilizing coded macroblocks residual information or motion vector information to determine whether to carry out block elimination filtering and when needs carry out block elimination filtering, selecting to adopt, thereby reduced computation complexity.The block-removal filtering method that the present invention adopts only determines whether to carry out block elimination filtering in the macro block one-level, and do not determine whether to carry out block elimination filtering in the piece one-level, because a macro block comprises a plurality of, determines whether that in the macro block one-level carrying out block elimination filtering greatly reduces computation complexity than determining whether to carry out block elimination filtering in the piece one-level.Simultaneously, because related sampled pixel point less and not overlapping in the block elimination filtering process, making the block elimination filtering process of vertical boundary of each macro block to walk abreast carries out, the block elimination filtering process of the horizontal boundary of each macro block also can walk abreast and carry out, therefore not only further reduce computation complexity, also reduced storage complexity simultaneously.The present invention is not limited to video coding or image code domain, can also extend in the application of other signal processing.
Description of drawings
Fig. 1 is existing deblocking filtering apparatus block diagram;
Fig. 2 is that existing block-removal filtering method carries out the related sampled pixel point schematic diagram of block elimination filtering to 4 * 4 levels or vertical boundary;
Fig. 3 is existing block-removal filtering method carries out block elimination filtering to block boundary a block boundary deblocking filtering apparatus block diagram;
Fig. 4 is a deblocking filtering apparatus block diagram of the present invention;
Fig. 5 is that block-removal filtering method of the present invention carries out the related sampled pixel point schematic diagram of block elimination filtering to 4 * 4 levels or vertical boundary;
Fig. 6 is block-removal filtering method of the present invention carries out block elimination filtering to block boundary a block boundary deblocking filtering apparatus block diagram.
Embodiment:
Accompanying drawing 1 shows the coding and decoding device of existing block-removal filtering method, comprises filtering strength decision device and the deblocking filtering apparatus that is used for macro block data is carried out block elimination filtering.Accompanying drawing 2 shows existing block-removal filtering method 4 * 4 levels or vertical boundary is carried out the related sampled pixel point of block elimination filtering.Wherein, thick lines represent to carry out the block boundary of block elimination filtering, p 0, p 1, p 2, p 3, q 0, q 1, q 2, q 3Each 4 sampled pixel point of expression block boundary both sides.Accompanying drawing 3 shows existing block-removal filtering method block boundary is carried out the block boundary deblocking filtering apparatus of block elimination filtering, p 0, p 1, p 2, p 3, q 0, q 1, q 2, q 3Expression is input to 8 sampled pixel points of block boundary deblocking filtering apparatus, P 0, P 1, P 2, Q 0, Q 1, Q 2Represent former sample pixel p 0, p 1, p 2, q 0, q 1, q 2Through the pixel value behind the block boundary block elimination filtering.
Codec is according to macro block (mb) type, and information such as macroblock coding residual sum motion vector are adjudicated current filtering strength, and according to filtering strength the input macro block data is carried out block elimination filtering and output.Existing block-removal filtering method needs each 4 sampled pixel point of block boundary both sides when 4 * 4 levels or vertical boundary are carried out block elimination filtering.
The weak point of existing method is that the needed information of judgement filtering strength is many, the condition judgment more complicated that relates to, and also related number of pixels is many at every turn when carrying out block elimination filtering, is difficult to parallel processing.Below at the deficiency of existing method, and present invention will be described in detail with reference to the accompanying.
Embodiment 1
Accompanying drawing 4 shows a kind of coding and decoding device that applies to video or image block elimination filtering, comprises the filter patterns decision device, is used for macro block data is carried out the deblocking filtering apparatus of block elimination filtering and is used for the TDMD time division multiplex device that timesharing is exported.The block-removal filtering method that accompanying drawing 5 shows the present invention's employing carries out the related sampled pixel point of block elimination filtering to 4 * 4 levels or vertical boundary.Wherein, thick lines represent to carry out the block boundary of block elimination filtering, p 0, p 1, q 0, q 1Each 2 sampled pixel point of expression block boundary both sides.Accompanying drawing 6 shows existing block-removal filtering method block boundary is carried out the block boundary deblocking filtering apparatus of block elimination filtering, p 0, p 1, q 0, q 1Expression is input to 4 sampled pixel points of block boundary deblocking filtering apparatus, P 0, P 1, Q 0, Q 1Represent former sample pixel p 0, p 1, q 0, q 1Through the pixel value behind the block boundary block elimination filtering.
Codec is according to macro block (mb) type, quantization parameter, and the quantization parameter threshold value is adjudicated current filter patterns, and according to filter patterns the input macro block data is carried out block elimination filtering and output.The block-removal filtering method that the present invention adopts needs each 2 sampled pixel point of block boundary two ends when 4 * 4 levels or vertical boundary are carried out block elimination filtering.
One, according to the quantization parameter CurrentQP of macro block (mb) type and macro block, determine the macro block filtering pattern by the following method:
(1) if current macro is an inter-coded macroblocks, then the macro block filtering pattern is the intra-frame macro block pattern;
(2) otherwise, if current macro is the skip mode coded macroblocks, and macroblock quantization parameter is less than a certain quantization parameter threshold value, then the macro block filtering pattern is filtered macroblock pattern not, this not filtering of macro block.Wherein, the quantization parameter threshold value equals quantization parameter threshold value default value or determines jointly that by quantization parameter threshold shift and quantization parameter threshold value default value the quantization parameter threshold shift can transmit, and also can transmit by other mode in encoding code stream;
(3) otherwise, the macro block filtering pattern is an inter macroblock.
Two, the derivation of block boundary threshold value
Accompanying drawing 4 expression piece p and piece q are at 4 sampled pixel points of level or vertical boundary both sides.Use P 0, P 1, Q 0And Q 1Represent sampled pixel point p respectively 0, p 1, q 0And q 1Carry out the pixel value behind the block elimination filtering.
If this macro block need carry out block elimination filtering, and following formula is true, then block boundary carried out block elimination filtering:
Abs(p 0-q 0)＜α
Function Abs is defined as: Abs ( x ) = x ; x > = 0 - x ; x < 0
Wherein α is the block boundary threshold value, can be according to the mean value CurrentQP of the quantization parameter CurrentQP of two pieces Av, and the skew AlphaCIOffset lookup table index IndexA of block boundary threshold value and intra-frame macro block filtering cutting coefficient index.
CurrentQP AvFor:
CurrentQP av＝(CurrentQP p+CurrentQP q+1)＞＞1
Index IndexA is:
IndexA＝Clip3(0，63，CurrentQP av+AlphaCIOffset)
Function C lip3 is defined as:
Figure C20051004900200072
According to the corresponding relation between index IndexA and threshold alpha, obtain the α value by table 1.
Table 1: the relation of block boundary threshold alpha and IndexA
Index α
0 0
1 0
2 0
3 0
4 0
5 0
6 1
7 1
8 1
9 1
10 1
11 2
12 2
13 2
14 3
15 3
Index α
16 4
17 4
18 5
19 5
20 6
21 7
22 8
23 9
24 10
25 11
26 12
27 13
28 15
29 16
30 18
31 20
Index α
32 22
33 24
34 26
35 28
36 30
37 33
38 33
39 35
40 35
41 36
42 37
43 37
44 39
45 39
46 42
47 44
Index α
48 46
49 48
50 50
51 52
52 53
53 54
54 55
55 56
56 57
57 58
58 59
59 60
60 61
61 62
62 63
63 64
Three, the block elimination filtering process of intra-frame macro block filter patterns
Sampled pixel point p to luminance block or both sides, chrominance block border 0, p 1, q 0And q 1Carry out block elimination filtering, the block elimination filtering process is as follows: at first according to the macro block filtering cutting coefficient of intra-frame macro block filter patterns correspondence, picture material obtains each pixel filter variable quantity that participates in the pixel correspondence of block elimination filtering process; Then, the pixel original value adds that the pixel filter variable quantity is the pixel value of block elimination filtering gained. 
    delta＝Clip3(-CI，CI，(((q0-p0)×4+(p1-q1)+4)＞＞3)    delta1＝delta＞＞1    if(p1＝＝p0)    {        P0＝Clip1(p0+delta)        P1＝Clip1(p1+delta1)    }    else        P0＝Clip1(p0+delta1)    if(q1＝＝q0)    {    
        Q0＝Clip1(q0-delta)        Q1＝Clip1(q1-delta1)    }    else        Q0＝Clip1(q0-delta1)
In the above-mentioned block elimination filtering process, delta, delta1 are respectively sampled pixel point p 0, q 0And p 1, q 1Corresponding filtering pixel increment.Function C lip1 is defined as: and Clip1 (x)=Clip3 (0,255, x).CI is called intra-frame macro block filtering cutting coefficient.Relation between CI and the IndexCI is referring to table 2.Wherein IndexCI is an intra-frame macro block filtering cutting coefficient index, is defined as follows:
IndexCI＝Clip3(0，63，QP av)
The relation of table 2:CI and IndexCI
IndexCI CI
0 0
1 0
2 0
3 0
4 0
5 0
6 0
7 0
8 0
9 0
10 0
11 0
12 0
13 0
14 0
15 0
IndexCI CI
16 0
17 0
18 0
19 0
20 0
21 0
22 1
23 1
24 1
25 1
26 1
27 1
28 1
29 2
30 2
31 2
IndexCI CI
32 2
33 2
34 2
35 3
36 4
37 4
38 4
39 4
40 5
41 5
42 5
43 5
44 5
45 6
46 6
47 6
IndexCI CI
48 6
49 7
50 7
51 8
52 9
53 9
54 10
55 11
56 12
57 13
58 14
59 15
60 16
61 16
62 16
63 16
Four, the block elimination filtering process of inter macroblocks filter patterns
Sampled pixel point p to luminance block or both sides, chrominance block border 0, p 1, q 0And q 1Carry out block elimination filtering, the block elimination filtering process is as follows: at first according to the macro block filtering cutting coefficient of inter macroblocks filter patterns correspondence, picture material obtains each pixel filter variable quantity that participates in the pixel correspondence of block elimination filtering process; Then, the pixel original value adds that the pixel filter variable quantity is the pixel value of block elimination filtering gained. 
    delta＝Clip3(-CP，CP，(((q0-p0)×4+(p1-q1)+4)＞＞3)    
    delta1＝delta＞＞1    if((p1＝＝p0)&&(Abs(p0-q0)＜((α＞＞2)+2)))    {        P0＝Clip1(p0+delta)        P1＝Clip1(p1+delta1)    }    else        P0＝Clip1(p0+delta1)    if((q1＝＝q0)&&(Abs(p0-q0)＜(α＞＞2)+2)))    {        Q0＝Clip1(q0-delta)        Q1＝Clip1(q1-delta1)    }    else        Q0＝Clip1(q0-delta1)
In the above-mentioned block elimination filtering process, CP is called inter macroblocks filtering cutting coefficient, and CP and CI can be derived mutually by formula and draw, and can be expressed as: (CI, CPOffset), wherein f () is specific function to CP=f, for example: CP=(CI＞＞1)+CPOffset
Wherein CPOffset is the skew of inter macroblocks filtering cutting coefficient.The skew of inter macroblocks filtering cutting coefficient can transmit in encoding code stream, also can transmit by other mode.
Embodiment 2
Accompanying drawing 4 shows a kind of coding and decoding device that applies to video or image block elimination filtering, comprises the filter patterns decision device, is used for macro block data is carried out the deblocking filtering apparatus of block elimination filtering and is used for the TDMD time division multiplex device that timesharing is exported.The block-removal filtering method that accompanying drawing 5 shows the present invention's employing carries out the related sampled pixel point of block elimination filtering to 4 * 4 levels or vertical boundary.Wherein, thick lines represent to carry out the block boundary of block elimination filtering, p 0, p 1, q 0, q 1Each 2 sampled pixel point of expression block boundary both sides.Accompanying drawing 6 shows existing block-removal filtering method block boundary is carried out the block boundary deblocking filtering apparatus of block elimination filtering, p 0, p 1, q 0, q 1Expression is input to 4 sampled pixel points of block boundary deblocking filtering apparatus, P 0, P 1, Q 0, Q 1Represent former sample pixel p 0, p 1, q 0, q 1Through the pixel value behind the block boundary block elimination filtering.
Codec is according to macro block (mb) type, quantization parameter, and the quantization parameter threshold value is adjudicated current filter patterns, and according to filter patterns the input macro block data is carried out block elimination filtering and output.The block-removal filtering method that the present invention adopts needs each 2 sampled pixel point of block boundary two ends when 4 * 4 levels or vertical boundary are carried out block elimination filtering.
One, according to the quantization parameter CurrentQP of macro block (mb) type and macro block, determine the macro block filtering pattern by the following method:
(1) if current macro is an inter-coded macroblocks, then the macro block filtering pattern is the intra-frame macro block pattern;
(2) otherwise, if current macro is the skip mode coded macroblocks, and macroblock quantization parameter is less than a certain quantization parameter threshold value, then the macro block filtering pattern is filtered macroblock pattern not, this not filtering of macro block.Wherein, the quantization parameter threshold value equals quantization parameter threshold value default value or determines jointly that by quantization parameter threshold shift and quantization parameter threshold value default value the quantization parameter threshold shift can transmit, and also can transmit by other mode in encoding code stream;
(3) otherwise, the macro block filtering pattern is an inter macroblock.
Two, the derivation of block boundary threshold value
Accompanying drawing 4 expression piece p and piece q are at 4 sampled pixel points of level or vertical boundary both sides.Use P 0, P 1, Q 0And Q 1Represent sampled pixel point p respectively 0, p 1, q 0And q 1Carry out the pixel value behind the block elimination filtering.
If this macro block need carry out block elimination filtering, and following formula is true, then block boundary carried out block elimination filtering:
Abs(p 0-q 0)＜α
Function Abs is defined as: Abs ( x ) = x ; x > = 0 - x ; x < 0
Wherein α is the block boundary threshold value, can be according to the mean value CurrentQP of the quantization parameter CurrentQP of two pieces Av, and the skew AlphaCIOffset lookup table index IndexA of block boundary threshold value and intra-frame macro block filtering cutting coefficient index.
CurrentQP AvFor:
CurrentQP av＝(CurrentQP p+CurrentQP q+1)＞＞1
Index IndexA is:
IndexA＝Clip3(0，63，CurrentQP av+AlphaCIOffset)
Function C lip3 is defined as:
Figure C20051004900200112
According to the corresponding relation between index IndexA and threshold alpha, obtain the α value by table 1.
Three, the block elimination filtering process of intra-frame macro block filter patterns
Sampled pixel point p to luminance block or both sides, chrominance block border 0, p 1, q 0And q 1Carry out block elimination filtering, the block elimination filtering process is as follows: at first according to the macro block filtering cutting coefficient of intra-frame macro block filter patterns correspondence, picture material obtains each pixel filter variable quantity that participates in the pixel correspondence of block elimination filtering process; Then, the pixel original value adds that the pixel filter variable quantity is the pixel value of block elimination filtering gained. 
    delta＝Clip3(-CI，CI，(((q0-p0)×4+(p1-q1)+4)＞＞3)    delta1＝delta＞＞1    if(p1＝＝p0)    {        P0＝Clip1(p0+delta)        P1＝Clip1(p1+delta1)    }    else        P0＝Clip1(p0+delta1)    if(q1＝＝q0)    {        Q0＝Clip1(q0-delta)        Q1＝Clip1(q1-delta1)    }    else        Q0＝Clip1(q0-delta1)
In the above-mentioned block elimination filtering process, delta, delta1 are respectively sampled pixel point p 0, q 0And p 1, q 1Corresponding filtering pixel increment.Function C lip1 is defined as: and Clip1 (x)=Clip3 (0,255, x).CI is called intra-frame macro block filtering cutting coefficient.Relation between CI and the IndexCI is referring to table 2.Wherein IndexCI is an intra-frame macro block filtering cutting coefficient index, is defined as follows:
IndexCI＝Clip3(0，63，QP av)
Four, the block elimination filtering process of inter macroblocks filter patterns
Sampled pixel point p to luminance block or both sides, chrominance block border 0, p 1, q 0And q 1Carry out block elimination filtering, the block elimination filtering process is as follows: at first according to the macro block filtering cutting coefficient of inter macroblocks filter patterns correspondence, picture material obtains each pixel filter variable quantity that participates in the pixel correspondence of block elimination filtering process; Then, the pixel original value adds that the pixel filter variable quantity is the pixel value of block elimination filtering gained. 
    delta＝Clip3(-CP，CP，(((q0-p0)×4+(p1-q1)+4)＞＞3)    delta1＝delta＞＞1    if((p1＝＝p0)&&(Abs(p0-q0)＜((α＞＞2)+2)))    
    {        P0＝Clip1(p0+delta)        P1＝Clip1(p1+delta1)    }    else        P0＝Clip1(p0+delta1)    if((q1＝＝q0)&&(Abs(p0-q0)＜(α＞＞2)+2)))    {        Q0＝Clip1(q0-delta)        Q1＝Clip1(q1-delta1)    }    else        Q0＝Clip1(q0-delta1)
In the above-mentioned block elimination filtering process, CP is called inter macroblocks filtering cutting coefficient.Relation between CP and the IndexCP is referring to table 3.Wherein IndexCP is an inter macroblocks filtering cutting coefficient index, is defined as follows: and IndexCP=Clip3 (0,63, QP Av).
The relation of table 3:CP and IndexCP
IndexCP CP
0 0
1 0
2 0
3 0
4 0
5 0
6 0
7 0
8 0
9 0
10 0
11 0
12 0
13 0
14 0
15 0
IndexCP CP
16 0
17 0
18 0
19 0
20 0
21 0
22 0
23 0
24 0
25 0
26 0
27 0
28 0
29 1
30 1
31 1
IndexCP CP
32 1
33 1
34 1
35 1
36 2
37 2
38 2
39 2
40 2
41 2
42 2
43 2
44 2
45 3
46 3
47 3
IndexCP CP
48 3
49 3
50 3
51 4
52 4
53 4
54 5
55 5
56 6
57 6
58 7
59 8
60 8
61 8
62 8
63 8
Embodiment 3
Accompanying drawing 4 shows a kind of coding and decoding device that applies to video or image block elimination filtering, comprises the filter patterns decision device, is used for macro block data is carried out the deblocking filtering apparatus of block elimination filtering and is used for the TDMD time division multiplex device that timesharing is exported.The block-removal filtering method that accompanying drawing 5 shows the present invention's employing carries out the related sampled pixel point of block elimination filtering to 4 * 4 levels or vertical boundary.Wherein, thick lines represent to carry out the block boundary of block elimination filtering, p 0, p 1, q 0, q 1Each 2 sampled pixel point of expression block boundary both sides.Accompanying drawing 6 shows existing block-removal filtering method block boundary is carried out the block boundary deblocking filtering apparatus of block elimination filtering, p 0, p 1, q 0, q 1Expression is input to 4 sampled pixel points of block boundary deblocking filtering apparatus, P 0, P 1, Q 0, Q 1Represent former sample pixel p 0, p 1, q 0, q 1Through the pixel value behind the block boundary block elimination filtering.
Codec is according to macro block (mb) type, quantization parameter, and the quantization parameter threshold value is adjudicated current filter patterns, and according to filter patterns the input macro block data is carried out block elimination filtering and output.The block-removal filtering method that the present invention adopts needs each 2 sampled pixel point of block boundary two ends when 4 * 4 levels or vertical boundary are carried out block elimination filtering.
One, according to the quantization parameter CurrentQP of macro block (mb) type and macro block, determine the macro block filtering pattern by the following method:
(1) if current macro is an inter-coded macroblocks, then the macro block filtering pattern is the intra-frame macro block pattern;
(2) otherwise, if current macro is the skip mode coded macroblocks, and macroblock quantization parameter is less than a certain quantization parameter threshold value, then the macro block filtering pattern is filtered macroblock pattern not, this not filtering of macro block.Wherein, the quantization parameter threshold value equals quantization parameter threshold value default value or determines jointly that by quantization parameter threshold shift and quantization parameter threshold value default value the quantization parameter threshold shift can transmit, and also can transmit by other mode in encoding code stream;
(3) otherwise, the macro block filtering pattern is an inter macroblock.
Two, the derivation of block boundary threshold value
Accompanying drawing 4 expression piece p and piece q are at 4 sampled pixel points of level or vertical boundary both sides.Use P 0, P 1, Q 0And Q 1Represent sampled pixel point p respectively 0, p 1, q 0And q 1Carry out the pixel value behind the block elimination filtering.
If this macro block need carry out block elimination filtering, and following formula is true, then block boundary carried out block elimination filtering:
f1(q 0，p 0，p 1，q 1)＜0
Wherein, f1 (x 0, x 0, x 1, x 1) function judged of certain conditions of expression.
Three, the block elimination filtering process of intra-frame macro block filter patterns
Sampled pixel point p to luminance block or both sides, chrominance block border 0, p 1, q 0And q 1Carry out block elimination filtering, the block elimination filtering process is as follows: at first according to the macro block filtering cutting coefficient of intra-frame macro block filter patterns correspondence, picture material obtains each pixel filter variable quantity that participates in the pixel correspondence of block elimination filtering process; Then, obtain the pixel value of block elimination filtering gained according to pixel original value and pixel filter variable quantity.
delta＝f2(CI，q 0，p 0，p 1，q 1)
P 0＝f3(p 0，delta)
P 1＝f4(p 1，delta)
Q 0＝f5(q 0，delta)
Q 1＝f6(q 1，delta1)
In the above-mentioned block elimination filtering process, f2 is a specific function, is used for obtaining the filter change amount, f3, and f4, f5, f6 are some specific functions, are used for to a p 0, p 1, q 0And q 1Carry out block elimination filtering.CI is called intra-frame macro block filtering cutting coefficient.Relation between CI and the IndexCI can be obtained by following formula: CI=f7 (IndexCI).
Wherein f7 is a specific function, is used for obtaining CI, and IndexCI is an intra-frame macro block filtering cutting coefficient index, is defined as follows:
IndexCI＝Clip3(0，63，QP av)。
Four, the block elimination filtering process of inter macroblocks filter patterns
Sampled pixel point p to luminance block or both sides, chrominance block border 0, p 1, q 0And q 1Carry out block elimination filtering, the block elimination filtering process is as follows: at first according to the macro block filtering cutting coefficient of inter macroblocks filter patterns correspondence, picture material obtains each pixel filter variable quantity that participates in the pixel correspondence of block elimination filtering process; Then, obtain the pixel value of block elimination filtering gained according to pixel original value and pixel filter variable quantity.
delta＝f2(CI，q 0，p 0，p 1，q 1)
P 0＝f3(p 0，delta)
P 1＝f4(p 1，delta)
Q 0＝f5(q 0，delta)
Q 1＝f6(q 1，delta1)
In the above-mentioned block elimination filtering process, f2 is a specific function, is used for obtaining the filter change amount, f3, and f4, f5, f6 are some specific functions, are used for to a p 0, p 1, q 0And q 1Carry out block elimination filtering.CP is called inter macroblocks filtering cutting coefficient.Relation between CP and the IndexCP can be obtained by following formula: CP=f7 (IndexCP).
Wherein f7 is a specific function, is used for obtaining CP, and IndexCP is an inter macroblocks filtering cutting coefficient index, is defined as follows:
IndexCP＝Clip3(0，63，QP av)。
The foregoing description is used for the present invention that explains, rather than limits the invention, and in the protection range of spirit of the present invention and claim, any modification and change to the present invention makes all fall into protection scope of the present invention.

Claims (10)

1. the method for video or image block elimination filtering is characterized in that,
(1) only utilize the quantization parameter of macro block (mb) type and macro block to determine whether to carry out block elimination filtering and the macro block filtering pattern of selection employing when needing block elimination filtering, and the macro block filtering pattern of not utilizing coded macroblocks residual information or motion vector information to determine whether to carry out block elimination filtering and when needs carry out block elimination filtering, selecting to adopt;
(2) determine the block elimination filtering process according to macro block filtering pattern and picture material self adaptation.
2. the method for block elimination filtering according to claim 1 is characterized in that, described (1) specifically is divided into following steps:
(a) quantization parameter according to macro block (mb) type and macro block judges whether macro block is carried out block elimination filtering and the macro block filtering pattern of selection employing when needs carry out block elimination filtering;
(b) pixel value according to block boundary threshold value and block boundary both sides determines whether block boundary is carried out block elimination filtering;
(c) obtain macro block filtering cutting coefficient according to macroblock quantization parameter and macro block filtering pattern;
(d) macro block filtering cutting coefficient and the picture material according to gained obtains the corresponding pixel filter variable quantity of each sampled pixel point that participates in the block elimination filtering process;
(e) obtain the resulting pixel value of block elimination filtering with original value and pixel filter variable quantity.
3. the method for block elimination filtering according to claim 2 is characterized in that, described step (a) is specially:
(i) if current macro is an inter-coded macroblocks, then the macro block filtering pattern is the intra-frame macro block pattern;
(ii) otherwise, if current macro is the skip mode coded macroblocks, and macroblock quantization parameter is less than a certain quantization parameter threshold value, then the macro block filtering pattern is filtered macroblock pattern not;
(iii) otherwise, the macro block filtering pattern is an inter macroblock.
4. the method for block elimination filtering according to claim 3 is characterized in that, described step (ii) in, described quantization parameter threshold value is quantization parameter threshold value default value or is determined jointly by quantization parameter threshold shift and quantization parameter threshold value default value.
5. the method for block elimination filtering according to claim 2 is characterized in that, in the described step (e), the resulting pixel value of described block elimination filtering is that original pixel value adds the pixel filter variable quantity.
6. the method for block elimination filtering according to claim 2, it is characterized in that, in the described step (c), described macro block filtering cutting coefficient comprises intra-frame macro block filtering cutting coefficient and inter macroblocks filtering cutting coefficient, described intra-frame macro block filtering cutting coefficient and inter macroblocks filtering cutting coefficient both can be can be derived by linear function mutually to obtain, and also can be separate.
7. the method for block elimination filtering according to claim 6, it is characterized in that, can derive obtain the time by linear function mutually when intra-frame macro block filtering cutting coefficient and inter macroblocks filtering cutting coefficient, comprise the skew of macro block filtering cutting coefficient in the linear function that concerns between expression intra-frame macro block filtering cutting coefficient and the inter macroblocks filtering cutting coefficient.
8. the method for block elimination filtering according to claim 2 is characterized in that, described step (d) also comprises, determines that according to picture material the block elimination filtering process will revise the number of the sampled pixel point of pixel value.
9. the method for block elimination filtering according to claim 2 is characterized in that, described step (d) comprises that also each corresponding pixel filter variable quantity of sampled pixel point that participates in the block elimination filtering process is obtained by formula mutually.
10. an application rights requires the coding and decoding device of the method for 1 described block elimination filtering, it is characterized in that it comprises:
One macro block filtering mode adjudging device, according to macro block (mb) type, quantization parameter, the quantization parameter threshold value is adjudicated current filter patterns, the filter patterns of decision current macro;
One deblocking filtering apparatus carries out block elimination filtering according to the output of filter patterns decision device to the current macro coded data and handles;
One time-division multiplexer;
Wherein, the filter patterns decision device links to each other with deblocking filtering apparatus, and deblocking filtering apparatus links to each other with TDMD time division multiplex device.
CNB2005100490020A 2005-01-31 2005-01-31 Deblocking filtering method and apparatus of video frequency or image Active CN100345450C (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CNB2005100490020A CN100345450C (en) 2005-01-31 2005-01-31 Deblocking filtering method and apparatus of video frequency or image

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CNB2005100490020A CN100345450C (en) 2005-01-31 2005-01-31 Deblocking filtering method and apparatus of video frequency or image

Publications (2)

Publication Number Publication Date
CN1652610A CN1652610A (en) 2005-08-10
CN100345450C true CN100345450C (en) 2007-10-24

Family

ID=34876495

Family Applications (1)

Application Number Title Priority Date Filing Date
CNB2005100490020A Active CN100345450C (en) 2005-01-31 2005-01-31 Deblocking filtering method and apparatus of video frequency or image

Country Status (1)

Country Link
CN (1) CN100345450C (en)

Families Citing this family (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8681867B2 (en) * 2005-10-18 2014-03-25 Qualcomm Incorporated Selective deblock filtering techniques for video coding based on motion compensation resulting in a coded block pattern value
US9628823B2 (en) * 2006-01-09 2017-04-18 Thomson Licensing Dtv Method and apparatus for providing reduced resolution update mode for multi-view video coding
CN101076124B (en) * 2006-05-18 2010-09-08 北京大学深圳研究生院 Filter for eliminating block
JP4747975B2 (en) * 2006-07-14 2011-08-17 ソニー株式会社 Image processing apparatus and method, program, and recording medium
US8401085B2 (en) 2006-09-07 2013-03-19 Lg Electronics Inc. Method and apparatus for decoding/encoding of a video signal
WO2008060125A1 (en) 2006-11-17 2008-05-22 Lg Electronics Inc. Method and apparatus for decoding/encoding a video signal
CN101395922A (en) * 2006-11-17 2009-03-25 Lg电子株式会社 Method and apparatus for decoding/encoding a video signal
KR101370287B1 (en) * 2006-11-22 2014-03-07 세종대학교산학협력단 Method and apparatus for deblocking filtering
CN101578874B (en) * 2007-01-04 2011-12-07 汤姆森特许公司 Methods and apparatus for reducing coding artifacts for illumination compensation and/or color compensation in multi-view coded video
EP2263381A1 (en) 2008-04-11 2010-12-22 Thomson Licensing Deblocking filtering for displaced intra prediction and template matching
CN101854539B (en) * 2009-04-03 2012-12-12 晨星软件研发(深圳)有限公司 Device and method for eliminating mosquito noise
EP2437500A4 (en) * 2009-05-29 2013-04-17 Mitsubishi Electric Corp Image encoding device, image decoding device, image encoding method, and image decoding method
CN101938647B (en) * 2009-06-30 2012-07-11 联咏科技股份有限公司 H264 deblocking-based multi-source filter and multi-source filtering method
JP5359657B2 (en) * 2009-07-31 2013-12-04 ソニー株式会社 Image encoding apparatus and method, recording medium, and program
WO2011096869A1 (en) * 2010-02-05 2011-08-11 Telefonaktiebolaget L M Ericsson (Publ) De-blocking filtering control
US8712173B2 (en) * 2010-03-12 2014-04-29 Mediatek Singapore Pte. Ltd. Methods for processing 2Nx2N block with N being positive integer greater than four under intra-prediction mode and related processing circuits thereof
PT3787289T (en) 2011-06-14 2022-09-30 Lg Electronics Inc Method for encoding and decoding image information
US9185404B2 (en) * 2011-10-07 2015-11-10 Qualcomm Incorporated Performing transform dependent de-blocking filtering
CN102521796A (en) * 2011-10-31 2012-06-27 南京大学 Image block effect removing method on basis of curvelet transformation
EP3767948A1 (en) * 2012-01-17 2021-01-20 Huawei Technologies Co., Ltd. In-loop filtering for lossless coding mode in high efficiency video coding
CN103220529B (en) * 2013-04-15 2016-02-24 北京大学 A kind of implementation method of coding and decoding video loop filtering
CN104284199B (en) * 2013-07-11 2019-02-01 Nxp股份有限公司 Video encoding/decoding method and device are carried out with the deblocking effect operation for reducing complexity
CN106911934B (en) * 2017-03-01 2020-03-03 北京奇艺世纪科技有限公司 Blocking effect removing method and deblocking filter
CN110290384A (en) * 2018-03-19 2019-09-27 华为技术有限公司 Image filtering method, device and Video Codec

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH08172628A (en) * 1994-12-20 1996-07-02 Victor Co Of Japan Ltd Reduction method of quantized noise caused when converted and coded image data are decoded and decoder for image data subjected to conversion coding
WO2003047262A2 (en) * 2001-11-29 2003-06-05 Matsushita Electric Industrial Co., Ltd. Coding distortion removal method and video encoding end decoding methods
WO2003094498A2 (en) * 2002-05-01 2003-11-13 Thomson Licensing S.A. Deblocking filter conditioned on pixel brightness

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH08172628A (en) * 1994-12-20 1996-07-02 Victor Co Of Japan Ltd Reduction method of quantized noise caused when converted and coded image data are decoded and decoder for image data subjected to conversion coding
WO2003047262A2 (en) * 2001-11-29 2003-06-05 Matsushita Electric Industrial Co., Ltd. Coding distortion removal method and video encoding end decoding methods
WO2003094498A2 (en) * 2002-05-01 2003-11-13 Thomson Licensing S.A. Deblocking filter conditioned on pixel brightness

Also Published As

Publication number Publication date
CN1652610A (en) 2005-08-10

Similar Documents

Publication Publication Date Title
CN100345450C (en) Deblocking filtering method and apparatus of video frequency or image
CN1224268C (en) Moving picture coding apparatus and moving picture decoding apparatus
CN1285216C (en) Image encoding method, image decoding method, image encoder, image decode, program, computer data signal, and image transmission system
CN1820512A (en) Method, medium, and filter removing a blocking effect
CN1867075A (en) Loop filtering method in image coding processing
CN101039432A (en) Method and apparatus for realizing self-adaptive quantization in coding process
CN101053259A (en) Fast intra mode prediction for a video encoder
CN101043621A (en) Self-adaptive interpolation process method and coding/decoding module
CN1543220A (en) Video decoding device and method
CN1287598C (en) Filtering method and apparatus for removing blocking artifacts and/or ringing noise
CN1777283A (en) Microblock based video signal coding/decoding method
CN1812580A (en) Deblocking control method considering intra bl mode and multilayer video encoder/decoder using the same
CN1960495A (en) Picture coding device, method, and program thereof
CN1767653A (en) Bit rate control method
CN1882083A (en) Method for reducing image blocking effect
CN1921625A (en) Filtering method and system of deblocking effect filter
CN1794816A (en) Moving picture compression encoding method, apparatus and program
CN1285214C (en) Loop filtering method and loop filter
CN1863319A (en) Fractional pixel interpolation method and encoding and decoding apparatus
CN1225919C (en) Image information encoding method and encoder, and image information decoding method decoder
CN101056408A (en) Method and apparatus for determining a threshold for skip mode in video coding
CN1224270C (en) Frame coding method of inter-frame coding frame for two stage predicting coding of macro block group structure
CN1756364A (en) Method for selecting intra-prediction mode
CN1266945C (en) Digital image processing method
CN1529509A (en) Video image sub-picture-element interpolation method and device

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
ASS Succession or assignment of patent right

Owner name: HUAWEI TECHNOLOGY CO., LTD.

Free format text: FORMER OWNER: ZHEJIANG UNIVERSITY

Effective date: 20080418

C41 Transfer of patent application or patent right or utility model
TR01 Transfer of patent right

Effective date of registration: 20080418

Address after: Bantian HUAWEI headquarters office building, Longgang District, Shenzhen

Patentee after: Huawei Technologies Co., Ltd.

Address before: Hangzhou City, Zhejiang Province, Xihu District Ancient Jade Road 20

Patentee before: Zhejiang University

TR01 Transfer of patent right
TR01 Transfer of patent right

Effective date of registration: 20211224

Address after: 450046 Floor 9, building 1, Zhengshang Boya Plaza, Longzihu wisdom Island, Zhengdong New Area, Zhengzhou City, Henan Province

Patentee after: Super fusion Digital Technology Co.,Ltd.

Address before: Bantian HUAWEI headquarters office building, Longgang District, Shenzhen

Patentee before: HUAWEI TECHNOLOGIES Co.,Ltd.