CN103428492A - Fast zig-zag scanning method for high-definition AVS coding - Google Patents

Fast zig-zag scanning method for high-definition AVS coding Download PDF

Info

Publication number
CN103428492A
CN103428492A CN201310297945XA CN201310297945A CN103428492A CN 103428492 A CN103428492 A CN 103428492A CN 201310297945X A CN201310297945X A CN 201310297945XA CN 201310297945 A CN201310297945 A CN 201310297945A CN 103428492 A CN103428492 A CN 103428492A
Authority
CN
China
Prior art keywords
zig
zag scanning
avs
coding
coefficient
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.)
Granted
Application number
CN201310297945XA
Other languages
Chinese (zh)
Other versions
CN103428492B (en
Inventor
王祖强
张贞雷
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.)
Shandong University
Original Assignee
Shandong University
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 Shandong University filed Critical Shandong University
Priority to CN201310297945.XA priority Critical patent/CN103428492B/en
Publication of CN103428492A publication Critical patent/CN103428492A/en
Application granted granted Critical
Publication of CN103428492B publication Critical patent/CN103428492B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Abstract

The invention relates to a fast zig-zag scanning method for high-definition AVS coding. The method includes the steps of 1, during AVS coding, according to a processing sequence of zig-zag scanning quantization coefficients, calculating the number p of nonzero coefficients in a quantization coefficient matrix during the process of inputting the quantization coefficient matrix to an AVS entropy coding module; 2, after inputting of the quantization coefficient matrix, subjecting the quantization coefficient matrix to zig-zag scanning, and calculating the number q of scanned nonzero coefficients; 3, judging the relation of p to q; when q is smaller than p, continuing zig-zag scanning for the quantization coefficient matrix; when q is equal to p, ending the zig-zag scanning for the quantization coefficient matrix; 4, after completion of the zig-zag scanning, continuing to perform AVS entropy coding. The method has the advantages that scanning tail zero coefficients is avoided so that zig-zag scanning speed is increased greatly; coding time is reduced greatly, and high-definition video coding can be achieved.

Description

A kind of method that in high definition AVS coding, zig-zag scans fast
Technical field:
The present invention relates to a kind of method that in high definition AVS coding, zig-zag scans fast, belong to multimedia technical field.
Background technology:
Along with the development of multimedia technology, people are more and more higher to the compression requirement of multimedia video, and new video encoding standard is arisen at the historic moment.AVS is the advanced video encoding standard of China's independent development, and the entropy coding, as the afterbody of whole AVS Video coding, occupies an important position in Video coding, and its coding efficiency can have influence on the operation of whole encoder.
Zig-zag scanning is an important part in the entropy coding, at first AVS prediction residual coefficient passes through Integer DCT Transform and quantification, the quantization parameter of quantizer output shows the rule that the low frequency coefficient amplitude is large, the high frequency coefficient amplitude is little, through the output after quantizing, it is a sparse matrix that comprises minority nonzero coefficient and a large amount of zero coefficient, be scanned into the rearrangement order by zig-zag, purpose is that nonzero coefficient is put together.For 8 * 8 coefficient matrixes, because the number of its nonzero coefficient is uncertain, in traditional way, be that 64 coefficients are carried out to whole scanning, comprise a large amount of zero coefficients in end end due to matrix, scanning for end end zero coefficient is nonsensical, therefore traditional scan method can take a large amount of clock cycle, and the speed of the coding that slowed down, be unfavorable for the high definition coding.Therefore how to accelerate zig-zag scanning, reduce insignificant sweep time, quite important for high definition AVS coding.
Summary of the invention:
For the deficiencies in the prior art, the invention provides a kind of method that in high definition AVS coding, zig-zag scans fast.The method can realize carrying out fast zig-zag scanning, greatly accelerates the speed of entropy coding, has also just greatly improved the speed of whole Video coding, is conducive to realize high definition AVS coding.
Technical scheme of the present invention is as follows:
A kind of method that in high definition AVS coding, zig-zag scans fast comprises that step is as follows:
(1) in the AVS coding, the prediction residual coefficient is at first according to prior art process Integer DCT Transform and quantification, a quantization parameter matrix that contains zero coefficient of quantizer output, before AVS entropy coding, above-mentioned quantization parameter matrix is reordered: the nonzero coefficient in the quantization parameter matrix is put together, in the AVS coding, adopt the quantization parameter processing sequence of zig-zag scanning, in to the input of AVS entropy coding module, quantizing the process of coefficient matrix, in the statistic quantification coefficient matrix, the number of nonzero coefficient, be designated as p; The present invention adopts the quantization parameter processing sequence of zig-zag scanning, and this scanning sequency is more easily followed the variation of non-zero quantized coefficients information;
(2) after completing the input of quantification coefficient matrix, the quantization parameter matrix is carried out to zig-zag scanning, in the process of scanning, the number of the nonzero coefficient that statistics scans, be designated as q;
(3) relation of judgement p and q:
In the time of q<p, continue the quantization parameter matrix is carried out to zig-zag scanning;
In the time of q=p, finish the zig-zag scanning to the quantization parameter matrix; Traditional zig-zag scanning will scan last coefficient of 8 * 8 quantization parameter matrixes always, scan 64 times, due in 8 * 8 general quantization parameter matrixes, p is much smaller than 64, therefore with traditional zig-zag scanning method, compare, utilize this quick zig-zag scanning method, can greatly add fast scan speed, also just accelerated the speed of coding simultaneously;
(4) after completing zig-zag scanning, proceed AVS entropy coding.
The invention has the advantages that:
Method of the present invention avoids scanning the zero coefficient of end section, has therefore greatly accelerated the speed of zig-zag scanning, and realize quick zig-zag scanning: in the quantization parameter matrix, the end end comprises 0 a large amount of coefficients.Due to improving constantly that the AVS field of video applications requires, high definition AVS Video coding has become main development trend, adopts this invention, can greatly reduce the scramble time, is conducive to realize the HD video coding.
The accompanying drawing explanation:
Fig. 1 is traditional zig-zag scanning;
Fig. 2 is realization flow figure of the present invention;
Fig. 3 is the implementation structure figure of the present invention's " nonzero coefficient statistical module ";
Fig. 4 is the deterministic process of the nonzero coefficient of the quick zig-zag scanning of the present invention;
Fig. 5 is the judge module structure chart of the nonzero coefficient that scans of zig-zag;
The effect schematic diagram that Fig. 6 is the quick zig-zag scanning of the present invention.
Embodiment:
Below in conjunction with accompanying drawing, the present invention is further illustrated, but be not limited to this.
Embodiment 1,
A kind of method that in high definition AVS coding, zig-zag scans fast comprises that step is as follows:
(1) in the AVS coding, the prediction residual coefficient is at first according to prior art process Integer DCT Transform and quantification, a quantization parameter matrix that contains zero coefficient of quantizer output, before AVS entropy coding, above-mentioned quantization parameter matrix is reordered: the nonzero coefficient in the quantization parameter matrix is put together, in the AVS coding, adopt the quantization parameter processing sequence of zig-zag scanning, in to the input of AVS entropy coding module, quantizing the process of coefficient matrix, in the statistic quantification coefficient matrix, the number of nonzero coefficient, be designated as p;
(2) after completing the input of quantification coefficient matrix, the quantization parameter matrix is carried out to zig-zag scanning, in the process of scanning, the number of the nonzero coefficient that statistics scans, be designated as q;
(3) relation of judgement p and q:
In the time of q<p, continue the quantization parameter matrix is carried out to zig-zag scanning;
In the time of q=p, finish the zig-zag scanning to the quantization parameter matrix;
(4) after completing zig-zag scanning, proceed AVS entropy coding.
As shown in Figure 1, in 8 * 8 quantization parameter matrixes after having quantized, 9 nonzero coefficients are arranged, for traditional zig-zag scanning, will be 64 whole been scanned of coefficient, if coefficient of 1 intermittent scanning needs 64 clock cycle could this quantization parameter matrix-scanning is complete.
As shown in Figure 2, according to realization flow of the present invention, in the input process of quantization parameter matrix, insert " nonzero coefficient statistical module ", after the quantization parameter input, just obtained the number of nonzero coefficient, be designated as variable p.
As shown in Figure 3, be the implementation structure figure of " nonzero coefficient statistical module ".Wherein cin is the quantization parameter of input.
As shown in Figure 4, be the deterministic process of the nonzero coefficient of the quick zig-zag of the present invention scanning, wherein cout is the quantization parameter scanned, the number that p is the nonzero coefficient that obtained by " nonzero coefficient statistical module " in input process.When zig-zag scans, the coefficient that scans out of judgement, send into " scanning judge module ", if non-zero, and q=q+1, if zero, q is constant.When q<p, scanning continues, and when q=p, finishes zig-zag scanning.
The judge module structure chart of the nonzero coefficient scanned for zig-zag as shown in Figure 5.
As shown in Figure 6, effect schematic diagram for the quick zig-zag scanning of the present invention, in Fig. 6, if each clock scan coefficient, only need 15 clocks just can finish zig-zag scanning, and traditional way needs 64 clocks, so the present invention has accelerated zig-zag scanning greatly, also just accelerate the speed of coding, be conducive to realize high definition AVS Video coding.

Claims (1)

1. the method for zig-zag scanning fast in high definition AVS coding comprises that step is as follows:
(1) in the AVS coding, the prediction residual coefficient is at first according to prior art process Integer DCT Transform and quantification, a quantization parameter matrix that contains zero coefficient of quantizer output, before AVS entropy coding, above-mentioned quantization parameter matrix is reordered: the nonzero coefficient in the quantization parameter matrix is put together, in the AVS coding, adopt the quantization parameter processing sequence of zig-zag scanning, in to the input of AVS entropy coding module, quantizing the process of coefficient matrix, in the statistic quantification coefficient matrix, the number of nonzero coefficient, be designated as p;
(2) after completing the input of quantification coefficient matrix, the quantization parameter matrix is carried out to zig-zag scanning, in the process of scanning, the number of the nonzero coefficient that statistics scans, be designated as q;
(3) relation of judgement p and q:
In the time of q<p, continue the quantization parameter matrix is carried out to zig-zag scanning;
In the time of q=p, finish the zig-zag scanning to the quantization parameter matrix;
(4) after completing zig-zag scanning, proceed AVS entropy coding.
CN201310297945.XA 2013-07-16 2013-07-16 The method of quickly zig-zag scanning in a kind of high definition AVS coding Expired - Fee Related CN103428492B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310297945.XA CN103428492B (en) 2013-07-16 2013-07-16 The method of quickly zig-zag scanning in a kind of high definition AVS coding

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310297945.XA CN103428492B (en) 2013-07-16 2013-07-16 The method of quickly zig-zag scanning in a kind of high definition AVS coding

Publications (2)

Publication Number Publication Date
CN103428492A true CN103428492A (en) 2013-12-04
CN103428492B CN103428492B (en) 2016-11-16

Family

ID=49652584

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310297945.XA Expired - Fee Related CN103428492B (en) 2013-07-16 2013-07-16 The method of quickly zig-zag scanning in a kind of high definition AVS coding

Country Status (1)

Country Link
CN (1) CN103428492B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH08153188A (en) * 1994-11-28 1996-06-11 Oki Electric Ind Co Ltd Zigzag scanning circuit
CN101198054A (en) * 2006-12-05 2008-06-11 华为技术有限公司 Variable length coding and decoding method and its coding and decoding device
CN101317462A (en) * 2005-11-30 2008-12-03 皇家飞利浦电子股份有限公司 Encoding method and apparatus applying coefficient reordering
CN101682771A (en) * 2007-06-15 2010-03-24 高通股份有限公司 adaptive coefficient scanning in video coding
CN102377994A (en) * 2010-08-05 2012-03-14 富士通株式会社 Context-adaptive variable-length coding method and system thereof
CN103004197A (en) * 2010-04-23 2013-03-27 吴秀美 Image encoding device and method

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH08153188A (en) * 1994-11-28 1996-06-11 Oki Electric Ind Co Ltd Zigzag scanning circuit
CN101317462A (en) * 2005-11-30 2008-12-03 皇家飞利浦电子股份有限公司 Encoding method and apparatus applying coefficient reordering
CN101198054A (en) * 2006-12-05 2008-06-11 华为技术有限公司 Variable length coding and decoding method and its coding and decoding device
CN101682771A (en) * 2007-06-15 2010-03-24 高通股份有限公司 adaptive coefficient scanning in video coding
CN103004197A (en) * 2010-04-23 2013-03-27 吴秀美 Image encoding device and method
CN102377994A (en) * 2010-08-05 2012-03-14 富士通株式会社 Context-adaptive variable-length coding method and system thereof

Also Published As

Publication number Publication date
CN103428492B (en) 2016-11-16

Similar Documents

Publication Publication Date Title
RU2543519C2 (en) Method and apparatus for image encoding and decoding using large transformation unit
TWI399929B (en) Cabac encoder and encoding method
CN103297777A (en) Method and device for increasing video encoding speed
CN1236624C (en) Quick full pixel movement evaluation method of multiple kinds of modules
US8908985B2 (en) Image processing including encoding information concerning the maximum number of significant digits having largest absolute value of coefficient data in groups
US20130101029A1 (en) Multimedia data encoding
CN104038765A (en) Rapid and efficient damage-free image compression method oriented to hardware achievement
CN1777286A (en) Image processing devices and methods
US20150063469A1 (en) Multipass encoder with heterogeneous codecs
CN103428492A (en) Fast zig-zag scanning method for high-definition AVS coding
CN103763561A (en) H264 video code parallel operation method
CN103533351A (en) Image compression method for multi-quantization table
CN109246437B (en) Image compression sensing method based on Reed-Solomon code
CN102215383A (en) Realization method of CAVLC (Context-Based Variable Length Coding) coder based on H.264
US8861880B2 (en) Image processing device and image processing method
CN108259896B (en) Columbus-Rice initial parameter self-adaptive decision method utilizing coefficient distribution characteristics
CN108124163B (en) All-zero block detection method suitable for HEVC (high efficiency video coding)
CN102625101B (en) Method for rapidly selecting H.264 interframe predication mode based on random pixel method
Petrovsky et al. Pipeline processing in real-time of CABAC decoder based on FPGA
CN101217665A (en) A parallel line scanning method of video frequency
CN101895757A (en) Method and system for reordering and inversely reordering predicted residual blocks
CN103414902A (en) AVC parallel coding method used for low power consumption applications
CN103533360B (en) A kind of AVS coding chip is optimized the method that code table stores
Xie et al. A real-time image row-compression method for high-definition USB cameras based on FPGA
So et al. Design and verification of intra prediction hardware for video streaming in IoT systems

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
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20161116

Termination date: 20180716

CF01 Termination of patent right due to non-payment of annual fee