CN100345449C - Method of entropy coding of transformation coefficient in image/video coding - Google Patents
Method of entropy coding of transformation coefficient in image/video coding Download PDFInfo
- Publication number
- CN100345449C CN100345449C CNB2005100552898A CN200510055289A CN100345449C CN 100345449 C CN100345449 C CN 100345449C CN B2005100552898 A CNB2005100552898 A CN B2005100552898A CN 200510055289 A CN200510055289 A CN 200510055289A CN 100345449 C CN100345449 C CN 100345449C
- Authority
- CN
- China
- Prior art keywords
- coefficient
- information
- value
- encoded
- nonzero
- 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.)
- Expired - Fee Related
Links
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
The present invention belongs to the field of image/video coding and decoding in signal processing, and relates to an entropy coding method with a transformation coefficient of the image/video coding and decoding. The entropy coding method of the present invention extracts the position information and the value information of non-zero coefficients respectively form a series of scanned coefficients so as to carry out the variable length coding; when coding the position information, the entropy coding method of the present invention combines continuous zero coefficients and the continuous non-zero coefficients into two-dimensional combined pairs of the run with the zero coefficients and the run with the non-zero coefficients and carries out the variable length coding on the combined pairs; when coding the value information, the entropy coding method of the present invention combines the value or the absolute value of the coefficients which continuously have the error of +1 or-1 and the subsequent absolute value larger than 1 into two-dimensional combination pairs of the run with the error of +1 or-1, and the value or the absolute value of the coefficients which have the absolute value larger than 1, and carries out the variable length coding on the combination pairs, or uses 1 bit to indicate whether all the absolute value of the series of the non-zero coefficients is 1. By further utilizing the position information and the statistical characteristics of the value information of the non-zero coefficients, the present invention has the compression performance more superior than that of the existing entropy coding method.
Description
Technical field
The invention belongs to the image/video encoding and decoding field in the signal processing, particularly in the image/video signals coding to the entropy coding of the coefficient after the scanning.
Background technology
In the existing image/video coding method, one images of video sequence generally is divided into several pieces and encodes, said be in the image as the pixel collection of coding base unit, be the pixel number more than or equal to the pixel collection on 1 square, rectangle or arbitrary shape border.Comprise prediction, conversion, quantification, scanning and entropy coding again as the step 1 of coding.Wherein, entropy coding method carries out variable-length encoding to this coefficient string after normally conversion coefficient being scanned into a coefficient string by certain rule again.
The entropy coding method of adopting in international video encoding standard MPEG-2 is that the coefficient string after will scanning earlier is combined into the two dimension combination of run and level one by one to (run, level), wherein run is meant the number (being called for short zero the distance of swimming) of continuous 0 before the nonzero coefficient, level is meant the positive and negative and absolute value of nonzero coefficient, again these two dimensions are made up carrying out variable-length encoding, be about to each (run, level) two dimension combination to all according to predefined two-dimentional code table in the standard be mapped to be one by several 0 and several 1 unique code words of translating of forming, and when the conversion coefficient end of a piece, show the end of this piece coefficient with a specific EOB code word.Use among the MPEG-2 (run, level) run of two dimension combination centering can regard the positional information of nonzero coefficient as, and level can regard the information of the value of nonzero coefficient as.Here said positional information is meant in the coefficient string after scanning the residing position of nonzero coefficient and when the coefficient string finishes, and the information of value is meant the absolute value and the sign of nonzero coefficient.MPEG-2 is that the information with positional information and value combines and encodes.
Entropy coding method at the conversion coefficient of up-to-date international video encoding standard in H.264/AVC mainly contains based on contextual adaptive variable length coding (CAVLC) with based on two kinds of contextual adaptive binary arithmetic codings (CABAC).CAVLC compares with CABAC lower complexity, has therefore only adopted CAVLC in basic framework (baseline profile) H.264/AVC.Continuous several 1 or-1 of the number of nonzero coefficient and end number is combined into coeff_token and carries out variable-length encoding based on adjacent block information in the coefficient string of CAVLC after scanning, value except that several nonzero coefficients 1 at end is represented with level_prefix and level_suffix, and carry out variable-length encoding, be remaining 0 that basic syntactic element is carried out based on contextual variable-length encoding with total_zeros and run_before.CAVLC than before the entropy coding method better utilization contextual information, so obtained than video compression standard more performance in the past.
However, also exist a lot of statistical property CAVLC not make full use of in the conversion coefficient, especially the correlation in the positional information of each nonzero coefficient, the correlation between the information of the value of each nonzero coefficient in addition.Therefore, existing entropy coding method is also imperfect at last.
Summary of the invention
The objective of the invention is in order to improve the performance of conventional images/video compression coding, a kind of new entropy coding method at conversion coefficient has been proposed, it carries out variable-length encoding respectively to the positional information of the nonzero coefficient in the coefficient string after scanning and the information of value, and more deep utilization the statistical property in the information of value of positional information and nonzero coefficient, as the combined characteristic of zero continuously number with the number of continuous nonzero coefficient, and whether the coefficient absolute value is complete 1 statistical property, or the like, thereby obtained the compression performance more more superior than existing entropy coding method.
One of technical scheme implementation method that the present invention proposes is: after conversion coefficient is scanned into a coefficient string by certain rule, this coefficient string is carried out variable-length encoding, it is characterized in that: the information of the value of the positional information of extraction nonzero coefficient and nonzero coefficient from this coefficient string, again these two kinds of information are carried out variable-length encoding respectively, when the positional information of this nonzero coefficient is encoded, it is right that the nonzero coefficient that continuous zero-sum is continuous is combined into the two dimension combination of the distance of swimming of zero the distance of swimming and nonzero coefficient, to described two dimension combination to carrying out variable-length encoding.
Two of the technical scheme implementation method that the present invention proposes is: after conversion coefficient is scanned into a coefficient string by certain rule, this coefficient string is carried out variable-length encoding, it is characterized in that: the information of the value of the positional information of extraction nonzero coefficient and nonzero coefficient from this coefficient string, again these two kinds of information are carried out variable-length encoding respectively, when the information of the value of this nonzero coefficient is encoded, with continuous 1 or-1 and following closely absolute value greater than the value or the absolute value of 1 coefficient be combined into ± 1 the distance of swimming and absolute value are right greater than the two dimension combination of the value of 1 coefficient or absolute value, to described two dimension combination to carrying out variable-length encoding.
Three of the technical scheme implementation method that the present invention proposes is: after conversion coefficient is scanned into a coefficient string by certain rule, this coefficient string is carried out variable-length encoding, it is characterized in that: the information of the value of the positional information of extraction nonzero coefficient and nonzero coefficient from this coefficient string, again these two kinds of information are carried out variable-length encoding respectively, when the information of the value of this nonzero coefficient is encoded, indicate with 1 bit whether the absolute value of this string nonzero coefficient is 1 entirely.
When carrying out variable-length encoding, can carry out variable-length encoding to the combination of the two dimension of above-mentioned each technical scheme according to the corresponding code table of the selection of the information self-adapting of having encoded.
Characteristics of the present invention and good result:
Entropy coding method in the image/video compressed encoding that the present invention proposes, it is encoded the information of the value of the positional information of the nonzero coefficient in the coefficient string after scanning and nonzero coefficient respectively, the information of various Itinerary Informations and various values has been carried out the better optimize combination, more helped variable-length encoding and improve compression performance.
The utilization that the present invention is more more deep than the entropy coding in existing image/video compressing and encoding method the statistical property in the information of value of positional information and nonzero coefficient, the more efficiently coding ± 1, thus the compression performance more more superior obtained than existing entropy coding method.
Embodiment
The entropy coding method embodiment of conversion coefficient is described in detail as follows in the image/video coding of the present invention:
Embodiment one: the coefficient string in the setting video coding after the change quantization scanning of certain piece is:
-1?-1?-3?3?-1?1?0?-4?0?0?0?2?-1?0?0?-2?0?0?0?1?-2?1?0?0?0?0?0?1?1?0?0?1
From this coefficient string, extract the information of the value of the positional information of nonzero coefficient and nonzero coefficient respectively, again these two kinds of information carried out variable-length encoding respectively:
1) positional information
Its combination of the distance of swimming (number of continuous nonzero coefficient) that is divided into several zero distances of swimming (continuous zero number) and nonzero coefficient is right, and as follows by the backward arrangement forward at last from this coefficient string:
(2，1)(5，2)(3，3)(2，1)(3，2)(1，1)(0，6)
Wherein, each bracket represents that a combination is right, and the first digit in the bracket is represented zero the distance of swimming, and second digit is represented the distance of swimming of the nonzero coefficient of zero back.For example, the coefficient string be 001 at last, it has only a continuous nonzero coefficient, have before it two continuous 0, so first combination is to being (2,1); Count forward, the coefficient string is 0000011 again, and therefore totally 50 and 2 continuous continuous nonzero coefficients make up being (5,2).By that analogy, up to the foremost-1-1-33-11 of coefficient string, totally 0 zero, 6 continuous nonzero coefficients, so last combination is to being (0,6).
Obtained zero the distance of swimming and nonzero coefficient the distance of swimming combination to after, need determine also when the positional information of this string coefficient finishes.In variable-length encoding, the methods that two kinds of indication end-of-encodes are arranged usually: a kind of is the right number of two dimension combination of the distance of swimming of zero the distance of swimming and nonzero coefficient of encoding in advance, another kind is after the two dimension combination of the distance of swimming of all zero distances of swimming and nonzero coefficient finishes to coding, in code stream, write a unique end code word of translating, come the end-of-encode of home position information.Except that these two kinds of methods, additive method can also be arranged.In the present embodiment, adopt first method, promptly the right number of two dimension combination of the distance of swimming of the distance of swimming by coding transmission zero and nonzero coefficient is indicated when to make up coding is finished.The right number of combination is 7 in the present embodiment.Like this, carry out variable-length encoding to making up right number 7 earlier, for example 7 be encoded to 010 in the present embodiment.
Successively above-mentioned each made up carrying out variable-length encoding afterwards.In video compression standard, preestablish one or more code table, coded combination to the time can come corresponding code table of adaptive selection and coding become corresponding unique code word of translating according to the information of having encoded.Here said information of having encoded can be the distance of swimming of the nonzero coefficient of the combination centering of having encoded, or the zero distance of swimming of all combination centerings of having encoded and the distance of swimming sum of nonzero coefficient, can also be other information that can from the content of having encoded, obtain.
In the present embodiment, be provided with four code tables in the video compression standard in advance for the positional information coding, and stipulate that in standard first combination is to adopting first code table, afterwards if the distance of swimming of the combination centering of just having encoded zero smaller or equal to 6, then use second code table, otherwise use the 3rd code table; Last combination is to using the 4th code table.Like this, first combination uses first code table to be encoded to 110 to (2,1), afterwards because just the distance of swimming of the combination centering of coding zero is less than 6, so second code table of (5,2) adaptive use is encoded to 001010; By that analogy, (3,3) also use second code table to be encoded to 00001, (2,1) also use second code table to be encoded to 01, (3,2) also use second code table to be encoded to 00010, (1,1) also uses second code table to be encoded to 10, last combination uses the 4th code table to be encoded to 000111 to (0,6).
Like this, the positional information of the nonzero coefficient of whole coefficient string is just encoded and is become 010 110 001,010 00,001 0,100,010 10 000111
2) Zhi information
Nonzero coefficient in the former coefficient string-1-1-3 3-1 1 0-4 000 2-1 0 0-2 000 1-2 10000011001 is extracted composition one row separately, promptly-1-1-3 3-1 1-4 2-1-2 1-2 1111
Continuous ± 1 at end in this nonzero coefficient string is all removed, and its be divided into ± 1 the distance of swimming (continuous 1 or-1 number) and absolute value following closely are right greater than the combination of the absolute value of 1 coefficient, promptly according to combination as described below:
(-1-1-3)(3)(-11-4)(2)(-1-2)(1-2)
With combinations thereof from after the distance of swimming of being arranged in one by one forward ± 1 and absolute value right, as follows greater than the combination of the absolute value of 1 coefficient:
(1，2)(1，2)(0，2)(2，4)(0，3)(2，3)
Here, each bracket represents that a combination is right, and the first digit in the bracket represents ± 1 the distance of swimming, and second digit represents that following closely absolute value is greater than the absolute value of 1 coefficient.For example in the nonzero coefficient string that extracts, last absolute value is-2 greater than 1 nonzero coefficient, and its front has only 1 continuous ± 1, so first combination is to being (1,2).Analogize forward according to this, up to the foremost of this nonzero coefficient string :-1-1-3, i.e. two continuous ± 1 ,-3 absolute values thereafter are 3, therefore combination is to being (2,3).
Successively these are made up carrying out variable-length encoding.In video compression standard, preestablish one or more code table, the combination of the information of these values of encoding to the time can come corresponding code table of adaptive selection and coding become corresponding unique code word of translating according to the information of having encoded.Here said information of having encoded can be the information of the absolute value of the combination centering of having encoded greater than the value of 1 coefficient, also can be the distance of swimming sum of combination centering ± 1 of having encoded, can also be other information that can obtain from the content of having encoded.In the end add a unique end code word of translating in addition and show when make up end-of-encode.
In the present embodiment, be provided with the code table of the information coding of three above-mentioned values of confession in the video compression standard in advance, and stipulate that in standard first combination is to adopting first code table, afterwards if the combination centering absolute value of just having encoded greater than the absolute value of 1 coefficient greater than 6, then use second code table, otherwise use the 3rd code table.Like this, first combination uses first code table to be encoded to 010 to (1,2), because less than 6, therefore (1,2) adopts the 3rd code table to be encoded to 010 greater than the absolute value of 1 coefficient for the combination centering absolute value of just having encoded, analogize in proper order, (0,2) adopts the 3rd code table to be encoded to 10, (2,4) adopt the 3rd code table to be encoded to 00101, (0,3) adopts the 3rd code table to be encoded to 011, (2,3) also adopt the 3rd code table to be encoded to 00010.Add one at last and finish code word 11, make up end-of-encode with expression.
The sign of nonzero coefficient is encoded to 1110101011010000 (1 representative is negative, and 0 representative just) in this coefficient string in addition.
Like this, the information of the value of the nonzero coefficient of whole coefficient string is just encoded and is become 010 010 10 00,101 01,100,010 11 1110101011010000
The bit that the bit that the positional information of nonzero coefficient is encoded out and the information of value are encoded out links together, thereby the coding result that obtains whole coefficient string is exactly 010 110 001,010 00,001 01 00,010 10 000,111 010,010 10 00,101 011 00,010 11 1110101011010000
Coding/decoding method is the inverse process of above-mentioned coding method.During decoding, this Bit String can uniquely decode the right number 7 of positional information combination successively, and the combination of the two dimension of 7 positional informations is to (2,1) (5,2) (3,3) (2,1) (3,2) (1,1) (0,6) be that the two dimension combination of information of value is to (1,2) (1 then,, 2) (0,2) (2,4) (0,3) sign of (2,3) and each nonzero coefficient--+-+-+--+-++ ++.According to the entropy coding method in the present embodiment, just these information can be recovered and the preceding consistent original coefficient string-1-1-3 3-1 1 0-4 000 2-1 0 0-2 00 0 1-2 10000011001 of encoding.
Embodiment two: the change quantization scanning coefficient string afterwards of establishing certain piece in the image encoding is:
0?-1?1?0?0?0?0?0?0?1?0?0?0?0?0?1?0?0?0?0?0?0?0?0?0?0?-1
From this coefficient string, extract the information of the value of the positional information of nonzero coefficient and nonzero coefficient respectively, again these two kinds of information carried out variable-length encoding respectively:
1) positional information.
The coding method of the positional information among the coding method of the positional information of the nonzero coefficient among the embodiment two and the embodiment one is identical.(present embodiment also can adopt other the existing methods among different and the embodiment one to come coding site information.)
Specifically: the combination of the distance of swimming that earlier it is divided into several zero distances of swimming and nonzero coefficient is right, and arranges as follows by backward forward at last from the coefficient string:
(10，1)(5，1)(6，1)(1，2)
Wherein, each bracket represents that a combination is right, and the first digit in the bracket is represented zero the distance of swimming, and second digit is represented the distance of swimming of the nonzero coefficient of zero back.For example, the coefficient string be the coefficient-1 of continuous 10 0 and 1 non-zeros at last, so first combination is to being (10,1); Count forward, the coefficient string is the coefficient 1 of continuous 50 and 1 non-zeros again, and therefore combination is to being (5,1).By that analogy, up to the foremost of coefficient string 0-11, totally 10,2 continuous nonzero coefficient, so last combination is to being (1,2).
For when the positional information of determining this string coefficient finishes, the two dimension combination that can count out the distance of swimming that has what zero distances of swimming and nonzero coefficient is right, and it is right to have 4 combinations in the present embodiment.
Like this, earlier carry out variable-length encoding to making up right number 4, for example present embodiment is encoded to 011, each combination of variable-length encoding is right successively then, be about to each combination to be encoded into a unique code word of translating according to the corresponding code table of setting up in the Standard of image compression, the combination of coding back to the time can be right according to the combination of having encoded information, come to select corresponding code table according to the rule of stipulating in the standard.
For example in compression standard, be provided with the code table of two coding site information in the present embodiment, just begun to adopt first code table, when zero the distance of swimming of a last coding less than 5 the time, the perhaps encoded all zero distances of swimming and the distance of swimming sum of nonzero coefficient were greater than 20 o'clock, adopt second code table, otherwise still adopt first code table.Like this, (10,1) adopt first code table to be encoded to 00000110, (5,1) still adopt first code table to be encoded to 001010, (6,1) also adopt first code table to be encoded to 00011, all zero distances of swimming that this moment is encoded and the distance of swimming sum of nonzero coefficient are 10+1+5+1+6+1=24, greater than 20, (thereby 1,2) adopts second code table to be encoded to 01.
Like this, the positional information of the nonzero coefficient of whole coefficient string is just encoded and is become 011 00,000,110 00,101,000,011 01
2) Zhi information.
In the present embodiment, the information of value adopt with embodiment one in different coding methods.Promptly adopt earlier a bit come the coded representation nonzero coefficient whether be entirely ± 1.If this bit is 1, then represent nonzero coefficient be entirely ± 1 if this bit is 0, then expression exists absolute value greater than 1 nonzero coefficient, still adopts the information coding method or the additive method of the value of embodiment one that the information of value is encoded afterwards.
In the present embodiment, nonzero coefficient is entirely ± 1, represent with a bit " 1 ".
Each encode afterwards ± 1 symbol, represent to bear, just represent with 0, promptly 10001 with 1
Like this, the information of the value of the nonzero coefficient of whole coefficient string is just encoded and is become 1 10001
The bit that the bit that the positional information of nonzero coefficient is encoded out and the information of value are encoded out links together, thereby the coding result that obtains whole coefficient string is exactly 011 00,000,110 001,010 00,011 01 1 10001
In the present embodiment, the information of encoded radio also can adopt the method identical with embodiment one, perhaps is defined in the method that adopts under certain condition among the embodiment one in standard, and adopts the method in the present embodiment under other conditions.Decision condition described here can be the number of all nonzero coefficients, or the right number of combination of the distance of swimming of encoded zero the distance of swimming and nonzero coefficient, or in the adjacent block ± and 1 number, perhaps other rules.But no matter use which kind of decision condition, all concrete regulation in advance.For example predesignate when the number of all nonzero coefficients smaller or equal to 6 the time, adopt 1 bit to represent whether the absolute value of these nonzero coefficients is 1 entirely, if be 1 entirely, then represent, otherwise represent with 0 with 1; When the number of all nonzero coefficients greater than 6 the time, adopt the coding method of the information of the value among the embodiment one.Like this, in the present embodiment, from distance of swimming of the nonzero coefficient of each two dimension combination centering of positional information number that obtains the nonzero coefficient the whole coefficient string of can suing for peace is 1+1+1+2=5, less than 6, and the absolute value of nonzero coefficient is 1 entirely, therefore still adopt a bit " 1 " come the coded representation nonzero coefficient be entirely ± 1.
Coding/decoding method is the inverse process of coding method.When decoding, the Bit String of encoding out in the present embodiment can uniquely decode the right number 4 of positional information combination successively, and the combination of the two dimension of 4 positional informations is to (10,1) (5,1) (6,1) (1,2), decode a bit " 1 " then, the absolute value of learning all nonzero coefficients is 1 entirely.Afterwards, decode each again ± 1 symbol-+++-.According to the entropy coding method in the present embodiment, just these information can be recovered and preceding consistent original coefficient string 0-1 10000001000001000000000 0-1 that encodes.
The above embodiment of the present invention only is used to content of the present invention is described, but does not limit the content that claim is protected.The applied field of the present invention is not limited to image or coding and decoding video field, can also extend in the application of other compressed encodings or signal processing.
Claims (5)
1, the entropy coding method in a kind of image/video compressed encoding, after conversion coefficient is scanned into a coefficient string by certain rule, this coefficient string is carried out variable-length encoding, it is characterized in that: the information of the value of the positional information of extraction nonzero coefficient and nonzero coefficient from this coefficient string, again these two kinds of information are carried out variable-length encoding respectively, when the positional information of this nonzero coefficient is encoded, it is right that the nonzero coefficient that continuous zero-sum is continuous is combined into the two dimension combination of the distance of swimming of zero the distance of swimming and nonzero coefficient, to described two dimension combination to carrying out variable-length encoding.
2, the entropy coding method in a kind of image/video compressed encoding as claimed in claim 1 is characterized in that: described two dimension combination when carrying out variable-length encoding, is carried out variable-length encoding according to the corresponding code table of the selection of the information self-adapting of having encoded.
3, entropy coding method in a kind of image/video compressed encoding, after conversion coefficient is scanned into a coefficient string by certain rule, this coefficient string is carried out variable-length encoding, it is characterized in that: the information of the value of the positional information of extraction nonzero coefficient and nonzero coefficient from this coefficient string, again these two kinds of information are carried out variable-length encoding respectively, when the information of the value of this nonzero coefficient is encoded, with continuous 1 or 1 and following closely absolute value greater than the value or the absolute value of 1 coefficient be combined into ± 1 the distance of swimming and absolute value are right greater than the two dimension combination of the value of 1 coefficient or absolute value, to described two dimension combination to carrying out variable-length encoding.
4, the entropy coding method in a kind of image/video compressed encoding as claimed in claim 3 is characterized in that: to described two dimension combination when carrying out variable-length encoding, carry out variable-length encoding according to the corresponding code table of the selection of the information self-adapting of having encoded.
5, the entropy coding method in a kind of image/video compressed encoding, after conversion coefficient is scanned into a coefficient string by certain rule, this coefficient string is carried out variable-length encoding, it is characterized in that: the information of the value of the positional information of extraction nonzero coefficient and nonzero coefficient from this coefficient string, again these two kinds of information are carried out variable-length encoding respectively, when the information of the value of this nonzero coefficient is encoded, indicate with 1 bit whether the absolute value of this string nonzero coefficient is 1 entirely.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100552898A CN100345449C (en) | 2005-03-18 | 2005-03-18 | Method of entropy coding of transformation coefficient in image/video coding |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100552898A CN100345449C (en) | 2005-03-18 | 2005-03-18 | Method of entropy coding of transformation coefficient in image/video coding |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1665303A CN1665303A (en) | 2005-09-07 |
| CN100345449C true CN100345449C (en) | 2007-10-24 |
Family
ID=35036160
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2005100552898A Expired - Fee Related CN100345449C (en) | 2005-03-18 | 2005-03-18 | Method of entropy coding of transformation coefficient in image/video coding |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100345449C (en) |
Families Citing this family (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100466742C (en) * | 2006-08-07 | 2009-03-04 | 清华大学 | Combined processing method for entropy decoding and converting flow line stage |
| KR20140101872A (en) * | 2007-01-11 | 2014-08-20 | 톰슨 라이센싱 | Methods and apparatus for using syntax for the coded_block_flag syntax element and the coded_block_pattern syntax element for the cavlc 4:4:4 intra, high 4:4:4 intra, and high 4:4:4 predictive profiles in mpeg-4 avc high level coding |
| CN101321283B (en) * | 2007-06-10 | 2010-04-07 | 华为技术有限公司 | Encoding/decoding method and device compatible with different block transformation |
| US8571104B2 (en) * | 2007-06-15 | 2013-10-29 | Qualcomm, Incorporated | Adaptive coefficient scanning in video coding |
| CN101944235B (en) * | 2009-09-18 | 2012-02-01 | 哈尔滨工程大学 | Image compression method based on fractional fourier transform |
| CN108777797B (en) | 2010-04-13 | 2021-01-29 | Ge视频压缩有限责任公司 | Device, method, storage medium and corresponding coding device for decoding a block of transform coefficients |
| CN102377994B (en) * | 2010-08-05 | 2014-05-07 | 富士通株式会社 | Context-adaptive variable-length coding method and system thereof |
| US9392301B2 (en) | 2011-07-01 | 2016-07-12 | Qualcomm Incorporated | Context adaptive entropy coding for non-square blocks in video coding |
| CN103002275B (en) * | 2011-09-13 | 2016-05-04 | 富士通株式会社 | For two coefficient scanning methods of CAVLC coding |
| CN102752592B (en) * | 2012-06-20 | 2015-02-11 | 深圳广晟信源技术有限公司 | Entropy coding method of video transformation coefficient |
| CN103346800B (en) * | 2013-07-08 | 2016-10-05 | 沈阳东软医疗系统有限公司 | A kind of data compression method and device |
| CN114554222B (en) * | 2017-07-07 | 2023-03-28 | 三星电子株式会社 | Video encoding method and device and video decoding method and device |
| CN117579846A (en) * | 2023-11-09 | 2024-02-20 | 四川新视创伟超高清科技有限公司 | Entropy coding method and device for lcevc |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH10191333A (en) * | 1996-12-27 | 1998-07-21 | Fuji Xerox Co Ltd | Image encoder |
| US6055272A (en) * | 1996-06-14 | 2000-04-25 | Daewoo Electronics Co., Ltd. | Run length encoder |
| JP2003264705A (en) * | 2002-03-11 | 2003-09-19 | Panasonic Communications Co Ltd | Image encoder and image encoding method |
-
2005
- 2005-03-18 CN CNB2005100552898A patent/CN100345449C/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6055272A (en) * | 1996-06-14 | 2000-04-25 | Daewoo Electronics Co., Ltd. | Run length encoder |
| JPH10191333A (en) * | 1996-12-27 | 1998-07-21 | Fuji Xerox Co Ltd | Image encoder |
| JP2003264705A (en) * | 2002-03-11 | 2003-09-19 | Panasonic Communications Co Ltd | Image encoder and image encoding method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1665303A (en) | 2005-09-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100345449C (en) | Method of entropy coding of transformation coefficient in image/video coding | |
| CN1215720C (en) | Preprocessing method applied to textures of arbitraily shaped objects | |
| CN100348051C (en) | An enhanced in-frame predictive mode coding method | |
| CN101185083B (en) | Hybrid variable length coding method and device for video compression | |
| AU2004240453B2 (en) | Combined runlength coding and variable length coding for video compression | |
| CN1148005C (en) | Quick joint image expert group Huffman coding and decoding method | |
| CN100488254C (en) | Entropy coding method and decoding method based on text | |
| RU2530341C1 (en) | Method and device for entropic coding/decoding of conversion factor | |
| CN1071526C (en) | Adaptive variable-length coding and decoding methods for image data | |
| CN1215647C (en) | Huffman decoder | |
| US7800520B2 (en) | Method and system for entropy coding | |
| CN101420614B (en) | Image compression method and device integrating hybrid coding and wordbook coding | |
| CN1527610A (en) | Entropy encoding method for encoding video predictive residual error coefficient | |
| CN101557517B (en) | Decoder, decoding method and apparatus | |
| CN1292594C (en) | Coding and decoding method and apparatus using plural scanning patterns | |
| CN1741616A (en) | Adaptive entropy coding/decoding method based on context | |
| CN1689026A (en) | Context-based adaptive variable length coding for adaptive block transforms | |
| CN1332522A (en) | Variable-length code using multiple mapping table and decoding method and apparatus | |
| CN1589023A (en) | Coding and decoding method and device for multiple coded list lengthening based on context | |
| CN1526240A (en) | Method and device for generating a scalable coded video signal from a non-scalable coded video signal | |
| CN101039374A (en) | Image lossless compression and image decompressing method | |
| CN1700771A (en) | Reinforced pixel domain code stream conversion method | |
| CN114143557B (en) | Low-complexity coding method for wavelet transformation high-frequency coefficient of video image | |
| CN1212014C (en) | Video coding method based on time-space domain correlation quick movement estimate | |
| CN1612613A (en) | Method and apparatus for motion estimation |
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 | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20071024 Termination date: 20120318 |