CN1816148A - Video-frequency data encoding method - Google Patents
Video-frequency data encoding method Download PDFInfo
- Publication number
- CN1816148A CN1816148A CN 200510005347 CN200510005347A CN1816148A CN 1816148 A CN1816148 A CN 1816148A CN 200510005347 CN200510005347 CN 200510005347 CN 200510005347 A CN200510005347 A CN 200510005347A CN 1816148 A CN1816148 A CN 1816148A
- Authority
- CN
- China
- Prior art keywords
- piece
- omission
- encoded
- block
- next line
- 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
Links
Images
Abstract
Using encoding mode prearranged in advance, the method encodes block of first row in each first frame of video data. The method also includes following steps: predicting ratio of omission block in next row to current encoding row; based on predicted result, determining whether ratio of omission block in next row is smaller than or equal to predefined ratio; if yes, when block of next row to be encoded being as omission block, then using code words corresponding to omission block as available encoding block type to carry out encoding operation; otherwise, when block of next row to be encoded not being as omission block, then using code words corresponding to sub block type of non-omission block to carry out encoding operation; otherwise, using standard encoding mode to encode block of next row to be encoded. Features are: improving encoding efficiency under high code rate, and not losing encoding efficiency under low code rate supplied by H.26L standard skip_run coding method.
Description
Technical field
The present invention relates to the video signal compression coding, particularly relate to a kind of coding method that is used for the video data of video coding system block type.
Background technology
As everyone knows, the main basis that can compress of multimedia video signal is: the first, exist a large amount of redundancys and this redundancy can recover after encoding and decoding on the vision signal undistortedly; The second, can utilize human vision property, under the condition that image change is not perceiveed, reduce the gray scale of quantized signal etc., exchange data compression for certain objective distortion.Multimedia data volume and amount of information relation can be expressed as I=D-du, and wherein I, D, du are respectively amount of information, data volume and amount of redundancy.Amount of information is the key data that will transmit, and amount of redundancy is exactly the data redundancy among the D, and data redundancy is useless data, therefore there is no need transmission.
The redundancy of multimedia video signal is present in structure and statistics two aspects.Structural redundancy shows as very strong space or in-frame correlation and time or frame-to-frame correlation.Generally speaking, most of regional signal of picture changes slowly, and especially background parts is almost constant.Therefore, vision signal is between neighbor, between adjacent lines, exist strong correlation between consecutive frame, and this correlation shows as spatial redundancy and time redundancy.On statistics,,, can think that still image is intact or enough good as long as the distortion of introducing during to image processing is difficult for discovering because human eye all has certain boundary to details resolution, Motion Resolution rate and the contrast resolution's of image sensation.
Can eliminate the effective method of frame-to-frame correlation at present is estimation, and promptly for the data of current image frame, at first search and its zone of mating most in former frame searches out the data that are complementary, and then both differences encoded.Therefore, the precision of estimation, promptly the matching degree of front and back two frame data is key factors of decision code efficiency.
Motion estimation algorithm has multiple, and wherein, block matching algorithm is simply effective because of its algorithm, and is easy to extensive integrated and be widely used in video coding.In block matching algorithm, every two field picture is divided into the piece (block) of the N * N pixel of two dimension, the row and column that every like this two field picture can divide the piece of serving as reasons to form, and line number is the pixel height of every two field picture and the ratio of N, columns is the pixel width of every two field picture and the ratio of N, and General N gets 16.Suppose that pixel in each piece all does the translational motion that equates, the piece of the N of present frame * N pixel searches with it the piece of coupling in the piece neighborhood window of previous frame correspondence, current detected is the motion vector that estimation obtains with the displacement of piece on two dimensional surface of mating most.
In standard H.26L, for each size is the piece of 16 * 16 pixels, the piece match pattern that block matching algorithm adopts has four kinds: 16 * 16 pixels, 16 * 8 pixels, 8 * 16 pixels and 8 * 8 pixel modes, the sub-block type that different piece match patterns is corresponding different.Block matching algorithm need adopt the size of this four seed block to carry out estimation respectively, chooses and the sub-block size of the current detected matching value minimum sub-block size as the coupling of this piece estimation.Then, to the different piece of piece that is detected and the sub-piece that mates most, just the image retention part is carried out discrete cosine transform, and the discrete cosine transform coefficient that obtains is carried out quantification treatment, and encodes.
At some in particular cases, if the sub-piece that mates most is the sub-piece of 16 * 16 pixels, and same position is arranged with the piece that is detected, the motion vector that this means this piece is zero, and the piece that is detected is very little with the image retention part of the sub-piece that mates most, so that the discrete cosine transform coefficient after the quantification of this piece is zero, this is called the omission piece.Omit piece and mean that this piece content does not almost change.This phenomenon often occurs in the background area of moving image.Accordingly, need be called encoding block by the piece that certain piece match pattern is encoded.
In the prior art, in standard H.26L, provide a kind of method of omitting block encoding efficient of can obviously improving under normal conditions: the skip_run coded system, be called for short SR (skip_run) scheme.This method adopts two code word Skip_Run and MB_Mode that block type information is encoded, wherein, omit the quantity of piece before the Skip_Run presentation code piece, the type of MB_Mode presentation code piece is provided with Skip_Run before the MB_Mode of each encoding block code word.The concrete coding codeword table that adopts can be referring to table 1.Wherein, the omission number of blocks before each encoding block all have one with the corresponding code word of omission number of blocks.For example, if 10 continuous omission pieces are arranged before a certain encoding block,, then can distribute the code word " 0001010 " of one 7 bit according to the codeword table shown in the table 1.The code word that the code word of encoding block type can the employing standard be provided with, concrete code word is not listed at table 1.
| Code word | Omit number of | Sub-block type | |
| 0 | 0 | 16×16 | |
| 1 | 1 | 8×8 | |
| 10 | 2 | 16×8 | |
| 11 | 3 | 8×16 | |
| ... | ... | ... |
Table 1
In general, the method for coding video data of used block type in the current H.26L standard, in the low range of code rates at the coding of normal dynamic or low dynamic video content near optimum state.But for the dynamically high of few of omission piece and high bit rate coding, the efficient of this coding method is again not enough.This mainly is because the percentage of omission number of blocks and targeted compression code check or quantification gradation are in close relations.For same video content, code check is low more, and it is many more to omit piece, and code check is high more, and it is few more to omit piece.Under situation worst, in the frame even may one omit piece and all do not have, if but adopt this method, still needing to put 1 bit and before each encoding block, equal 0 with expression Skip_Run, this is a kind of waste really.Along with increasing application request is encoded with high code check, this method can not satisfy the requirement of high coding efficiency.
Summary of the invention
In view of this, main purpose of the present invention is to provide a kind of coding method of video data, can improve the code efficiency under the high code check situation.
For achieving the above object, technical scheme of the present invention is achieved in that
The invention provides a kind of coding method of video data, this method utilizes predefined coded system that the piece of each frame video data first row is encoded; This method is further comprising the steps of:
A. predict the omission piece ratio of present encoding row next line piece in this frame;
B. according to predicting the outcome of obtaining among the step a, whether the omission piece ratio of judging the next line piece of being predicted is less than or equal to estimated rate, if, then the piece that is encoded at next line is to utilize when omitting piece to omit piece as the pairing omission block type of codified block type code word, when the piece that is encoded of next line is non-omission piece, utilize the pairing code word of sub-block type of non-omission piece, each piece that is encoded of next line is encoded; If not, then utilize the standard code mode that the piece that is encoded of next line is encoded.
Described predefined coded system is standard skip_run coded system H.26L.
The method that the piece ratio is omitted in the described prediction of step a is: determine the omission piece in the present encoding row, omit the number of piece in the statistics present encoding row, the number of the omission piece added up divided by the piece sum in the present encoding row, is obtained omitting the omission piece ratio of piece ratio as the next line piece.
In a frame the quantity of code line greater than 1 o'clock, the method that the piece ratio is omitted in the described prediction of step a is: determine before the present encoding row adjacent N omission piece capable, add up the omission piece number of described N in capable, with the number of the omission piece added up divided by described N capable in the sum of piece, obtain omitting the omission piece ratio of piece ratio as the next line piece, described N is less than or equal to the quantity of described code line and greater than 1 natural number.
Wherein, the method for described definite omission piece is: obtain the motion vector of piece, and obtain the compressed symbolic of this piece, if the motion vector of this piece that is obtained and compressed symbolic are 0, determine that then this piece is the omission piece, otherwise determine that this piece is not to omit piece.
Described estimated rate is 10%～40%.Preferably, described estimated rate is 25%.
Preferably, described size is 16 * 16 pixels.And preferably, described omission block type code word is 010; Described sub-block type is 16 * 16 pixels, and corresponding code word is 1; Described sub-block type is 8 * 8 pixels, and corresponding code word is 011; Described sub-block type is 16 * 8 pixels, and corresponding code word is 00100; Described sub-block type is 8 * 16 pixels, and corresponding code word is 00101.
The described current piece that is encoded is not that further comprise: compressed symbolic and motion vector information to this piece are encoded when omitting piece.
In the such scheme, further comprise decode procedure, specifically comprise:
Utilize the pairing decoding process of coded system of each frame video data first row being encoded that first row of each frame video data is decoded with described cataloged procedure;
Predict the omission piece ratio in the next line video data of current decoded row in this frame;
According to predicting the outcome of obtaining, judge whether the omission piece ratio in the next line of being predicted is less than or equal to estimated rate, if, then adopt in the described cataloged procedure to be less than or equal to code word identical under the estimated rate situation, omission piece and encoding block in the next line video data are decoded at described omission piece ratio; If not, then adopt the decoding process of the standard code mode correspondence described in the cataloged procedure that the next line video data is decoded.
Preferably, described standard code mode is standard skip_run coded system H.26L.
Wherein, the estimated rate in the described decode procedure equals the estimated rate in the described cataloged procedure.
The piece that described current decoding obtains is not that further comprise: compressed symbolic and motion vector information to this piece are decoded when omitting piece.
Described predefined coded system can also for: whether according to the piece that is encoded is to omit piece, utilizes and omits piece as the pairing omission block type of codified block type code word, or the pairing code word of sub-block type of non-omission piece, and piece is encoded to being encoded.
The coding method of the video data of block type provided by the present invention utilizes predefined coded system that the piece of first row of each frame video data is encoded; Predict the omission piece ratio of the next line piece of present encoding row in this frame then; When the omission piece ratio that predicts the next line piece is less than or equal to estimated rate, whether according to the piece that is encoded is to omit piece, utilize and omit piece as the pairing omission block type of codified block type code word, or the pairing code word of sub-block type of non-omission piece, each piece that is encoded of next line is encoded.That is to say, when not omitting piece before the encoding block, needn't export specific code word and represent not omit piece that like this, the inventive method can improve that to omit piece be code efficiency under the high code check situation less.
Simultaneously, the present invention is at the omission piece ratio of next line during greater than estimated rate, utilizes H.26L standard skip_run coded system that the piece that is encoded of next line is encoded.Like this, the code efficiency under the many promptly low code check conditions of the omission piece ratio that provides in the current H.26L standard is not provided in this coding method.Therefore, the inventive method is a kind of coding method that can handle astable type data source simply.And, code word in omitting the coding codeword table that adopts when the piece ratio is less than or equal to estimated rate is to set according to the probability of the sub-block type appearance of various encoding blocks, the most sub-block type correspondence of frequent appearance is set to the minimum code word of bit, and this has just further guaranteed the efficient of coding.
Description of drawings
Fig. 1 realizes the method flow diagram of the video data encoding of block type for the present invention;
Fig. 2 realizes the method flow diagram of the video data encoding of 16 * 16 pixel block types for one embodiment of the invention;
Fig. 3 realizes the method flow diagram of the video data decoding of 16 * 16 pixel block types for one embodiment of the invention.
Embodiment
The present invention is further described in more detail below in conjunction with drawings and the specific embodiments.
Main thought of the present invention is: utilize predefined coded system that the piece of first row of each frame video data is encoded; Predict the omission piece ratio of the next line piece of present encoding row in this frame; According to predicting the outcome, whether the omission piece ratio of judging next line is less than or equal to estimated rate, if, then the piece that is encoded at next line is to utilize when omitting piece to omit piece as the pairing omission block type of codified block type code word, when the piece that is encoded of next line is non-omission piece, utilize the pairing code word of sub-block type of non-omission piece, each piece that is encoded of next line is encoded; If not, then utilize the standard code mode that the piece that is encoded of next line is encoded.Wherein, described standard code mode is generally H.26L standard skip_run coded system.
Particularly,, the present invention proposes a new variable length code word table, be used for comprehensively expressing the block encoding schema type information for omitting the lower situation of percentage that piece occurs.Suppose that every two field picture is divided into the piece of 16 * 16 pixels of two dimension, the piece match pattern that is adopted during coupling has four kinds: 16 * 16 pixels, 16 * 8 pixels, 8 * 16 pixels, 8 * 8 pixels, the sub-block type of then various match pattern correspondences and the pairing code word of omission piece are as shown in table 2.
| Code word | |
| 1 | 16×16 |
| 010 | Omit piece |
| 011 | 8×8 |
| 00100 | 16×8 |
| 00101 | 8×16 |
Table 2
As seen, this codeword table will be omitted piece and encode as a seed block type.This codeword table is set the corresponding codes code word according to the probability that each seed block type occurs.Because the sub-block type of 16 * 16 pixel modes is sub-block types of the most frequent appearance, and therefore the minimum code word of bit " 1 " expression is set.Adopt code word " 010 " to represent to omit piece, adopt the sub-block type of 011 expression, 8 * 8 pixel modes, adopt the sub-block type of 00100 expression, 16 * 8 pixel modes, adopt the sub-block type of 00101 expression, 8 * 16 pixel modes.Like this, the encoding block that adopts certain piece match pattern coding if desired is that non-omission piece does not omit piece before, then need not to add the code word of omitting block type.The scheme of the coding of codeword table shown in the employing table 2 is called comprehensive elongated degree (CVL, combined variable length) encoding scheme.
As shown in Figure 1, this method may further comprise the steps:
Step 2, according to the omission piece ratio in the video data of having encoded in the frame, predict the omission piece ratio in the next line piece in this frame.Wherein, omission piece ratio is meant and omits the percentage that piece accounts for these all pieces of row in delegation's piece.
To the prediction of omitting the piece ratio in the next line video data multiple mode can be arranged, comprise: the omission piece to N before next line adjacent in capable is added up, and wherein N is the natural number that is less than or equal to the described quantity of code line.Be exactly only the omission piece in the lastrow data to be added up when N=1, in N equals this frame the front all during the quantity of code line, exactly to all the omission piece in the code line piece add up, when N is between above-mentioned two numerical value, exactly the omission piece in adjacent several capable pieces is added up; Then with statistics, as the probability of the omission piece in the next line data.According to motion relevance, omit piece and video content and be closely related, video content is continually varying basically, thereby the omission piece also is a continually varying.Therefore, can to handle the distribution probability of capable omission piece by the distribution probability prediction of the omission of the adjacent lines in two field picture piece.
Step 3～5, according to the statistical forecast result in the step 2, judge whether the omission piece ratio of being predicted is less than or equal to estimated rate, if, then the piece that is encoded to next line uses the CVL encoding scheme, promptly the piece that is encoded at next line is to utilize when omitting piece to omit piece as the pairing omission block type of codified block type code word, when the piece that is encoded of next line is non-omission piece, utilize the pairing code word of sub-block type of non-omission piece, the piece that is encoded of next line is encoded; If not, then the piece that is encoded to next line is provided by the SR encoding scheme that provides in the existing standard, promptly adds the code word that the quantity of piece is omitted in expression before each expression needs to adopt the code word of sub-block type of encoding block of certain piece match pattern coding.Usually, the value of estimated rate is 10%～40%.
Realize the process of Code And Decode of the present invention below with concrete example explanation.Among the following embodiment, every two field picture is divided into the piece of 16 * 16 pixels of two dimension, and the sub-block type that coupling obtains has: 16 * 16 pixels, 16 * 8 pixels, 8 * 16 pixels, 8 * 8 pixels.Because the piece of 16 * 16 pixels is commonly referred to macro block, so in the present embodiment, will omit piece and correspondingly be called the omission macro block, and encoding block correspondingly is called coded macroblocks.
Fig. 1 is the method flow diagram of the video data encoding of one embodiment of the invention realization 16 * 16 pixel block types, may further comprise the steps:
The value of step 100, variable that initialization is necessary is 0.
Wherein variable comprises: I, J, SKIP_RUN and SMBN.The index of macroblock number during I and J are respectively row and are listed as, 0≤I＜line number, line number=vertical frame dimension degree/16,0≤J＜columns, columns=frame width/16; SKIP_RUN represents the preceding quantity of omitting macro block of a coded macroblocks; SMBN represents to omit in the delegation quantity of macro block.Encoder need be encoded successively to the macro block of each frame video image data, and each macro block is only encoded once.According to the index of row with the row number of macroblocks, circulation is encoded to intra-frame macro block, and current line is listed as circulation, and the quantity of utilizing SKIP_RUN to add up the preceding omission macro block of a certain coded macroblocks is with the sum that omits macro block in the SMBN statistics current line.
In the present embodiment, FLAG represents the current encoding scheme that adopts with sign, establishes FLAG=0 and represents that the encoding scheme of current employing is the SR encoding scheme, establishes FLAG=1 and represents that the encoding scheme of current employing is the CVL encoding scheme.In the present embodiment, preferably, initial setting up FLAG=0 just adopts standard SR encoding scheme that the first row macro block is encoded.Certainly, also can adopt CVL encoding scheme of the present invention that the first row macro block is encoded.Correspondingly, decode procedure need adopt the decoding scheme of selected encoding scheme correspondence to the processing of first row.
Step 101, whether judge I less than line number, if then execution in step 102; An otherwise frame end.
Step 102, whether judge J less than columns, if then execution in step 103; Otherwise execution in step 114.
Step 103, obtain compressed symbolic, promptly obtain residual fraction, carry out discrete cosine transform and quantification, the discrete cosine transform coefficient after obtaining quantizing when the macro block of pre-treatment by motion search or intra-prediction.
Step 104～105, according to the discrete cosine transform coefficient after the quantification that is obtained, judge whether the macro block when pre-treatment is to omit macro block, if, then add up the omission macro block in this row macro block, the value that is about to current SMBN adds 1, continues execution in step 106; If not, then execution in step 109.
In the present embodiment,, the omission macro block ratio in the next line is predicted by only the omission macro block in the present encoding row being carried out statistical.This step is exactly to be used for the omission macro block of present encoding row is added up.
Wherein, whether the inquiry current macro is that the method for omitting macro block is: judge whether the discrete cosine coefficient after motion vector and the quantification is 0, if then current macro is to omit macro block, otherwise current macro is a coding module.Wherein, motion vector can obtain according to the block matching algorithm computing before encoder is encoded.
Step 106～108, judge whether FLAG is 1, if, then utilize table 2 output to omit the code word of macro block, promptly export code word " 010 ", otherwise the value of current SKIP_RUN is added 1; Continue execution in step 113.
Step 109～112, judge whether FLAG is 0, if, then utilize the code word in the table 1 that the sub-block type information of omission macro block before the current macro and coded macroblocks is encoded, and SKIP_RUN is set is 0, if not, then utilize the code word in the table 2 that the sub-block type information of current macro is encoded; Continue execution in step 113.
Wherein, utilizing the code word in the table 1 is exactly that the value of current SKIP_RUN is encoded to the omission macro block before the current macro.
Step 113, encode according to encryption algorithm pair all out of Memory relevant H.26L with coded macroblocks.Out of Memory comprises the discrete cosine transform coefficient after motion vector, the quantification etc.
Step 114, the value of current J is added 1, turn back to step 102.
This step is used to be listed as circulation, so that all macro blocks in the delegation are encoded.
Step 115～117, judge whether the SMBN and the ratio of width are less than or equal to estimated rate, if FLAG=1 then is set, otherwise FLAG=0 is set; Continue execution in step 118.
This step is used for being identified for the encoding scheme of next line after delegation's macroblock coding is finished.Be less than or equal to estimated rate if omit the ratio that macro block accounts for the full line macro block in the current line, such as 25%, promptly SMBN/NMB≤25% is then selected the CVL encoding scheme, otherwise selects the SR encoding scheme.
Step 118, J=0 is set, the value of current I is added 1, turn back to step 101.
This step is used to go circulation, each is gone coding video data.
Coding method provided by the invention only needs very little extra computation amount and memory usage, can ignore in implementation procedure.
The present invention provides the decoding process corresponding to above-mentioned coding method simultaneously, as shown in Figure 2, may further comprise the steps:
The value of step 200, variable that initialization is necessary is 0.Wherein variable comprises: current line number I and current columns J, be used for representing that delegation omits the SMBN of the quantity of macro block, and the SKIP_RUN of the quantity of the omission macro block before the coded macroblocks that obtains that is used to represent to decode.
Decoder need be decoded successively to the macro block of each frame video image data, and each macro block is only decoded once.According to the index of row with the row number of macroblocks, circulation is decoded to intra-frame macro block, rank of advanced units circulation.In the present embodiment, same represent the current decoding scheme that adopts with identifying FLAG, establishing the decoding scheme that FLAG=0 represents current employing is the SR scheme, and establishing the decoding scheme that FLAG=1 represents current employing is the CVL scheme.In the present embodiment, initial setting up FLAG=0, just first row adopts the SR scheme to decode.
Step 203～204, judge whether FLAG is 1, if, utilize table 2 codeword decoding macro block (mb) type, continue execution in step 205; Otherwise execution in step 208.
Whether the macro block that pre-treatment is worked as in step 205～207, judgement is to omit macro block, if then add up the omission macro block in these frame data, the value that is about to current SMBN adds 1, and carry out other processing to omitting macro block, if not, then the out of Memory of current macro is decoded; Continue execution in step 213.
In this step, according to table 2, if the code word that decoding obtains is " 010 ", then current macro is to omit macro block, and at this moment decoder adds 1 with SMBN, and carries out other necessary processing to omitting macro block, such as duplicate existing corresponding video content from last frame data.In addition, not to omit macro block but during coded macroblocks in current macro, all out of Memory of this coded macroblocks that also needs to decode are such as motion vector and discrete cosine transform coefficient etc.Belong to existing necessary the processing to omitting other necessary processing that macro block carried out, do not repeat them here.
Codeword decoding SKIP_RUN in step 208, the use table 1 obtains the preceding value of omitting the quantity SKIP_RUN of macro block of a coded macroblocks, and carries out other necessary processing to omitting macro block.
This step mainly is to upgrade row J and SMBN by the value of SKIP_RUN.
Step 211～212, utilize code word in the table 1, the current macro out of Memory is decoded the sub-block type information decoding of current coding macro block; Continue execution in step 213.
This step is used to be listed as circulation, so that all macro blocks in the delegation are decoded.
Step 214～216, judge whether the SMBN and the ratio of width are less than or equal to estimated rate, if FLAG=1 then is set, otherwise FLAG=0 is set; Continue execution in step 217.
This step is used for after the decoding of delegation macro block is finished, and is identified for the decoding scheme of next line according to the rule identical with encoder-side, and the set estimated rate of the estimated rate that is provided with and cataloged procedure is identical.Particularly, be less than or equal to estimated rate if omit the ratio that macro block accounts for the full line macro block in the current line, such as cataloged procedure be provided with 25%, promptly SMBN/NMB≤25% is then selected the CVL encoding scheme, otherwise selects the SR encoding scheme.
This step is used to go circulation, so that the next line video data is decoded.
As can be seen, coding method provided by the invention can improve the code efficiency under the high code check situation, simultaneously, code efficiency under the low code check condition that this coding method is not lost current H.26L standard method and provided, therefore, the inventive method is a kind of coding method that can handle astable type data source simply.
In a word, the above only is preferred embodiment of the present invention, and is in order to restriction the present invention, within the spirit and principles in the present invention not all, any modification of being done, is equal to replacement, improvement etc., all should be included within protection scope of the present invention.
Claims (15)
1, a kind of coding method of video data is characterized in that, this method utilizes predefined coded system that the piece of each frame video data first row is encoded; This method is further comprising the steps of:
A. predict the omission piece ratio of present encoding row next line piece in this frame;
B. according to predicting the outcome of obtaining among the step a, whether the omission piece ratio of judging the next line piece of being predicted is less than or equal to estimated rate, if, then the piece that is encoded at next line is to utilize when omitting piece to omit piece as the pairing omission block type of codified block type code word, when the piece that is encoded of next line is non-omission piece, utilize the pairing code word of sub-block type of non-omission piece, each piece that is encoded of next line is encoded; If not, then utilize the standard code mode that the piece that is encoded of next line is encoded.
2, coding method according to claim 1 is characterized in that, described predefined coded system is standard skip_run coded system H.26L.
3, coding method according to claim 1, it is characterized in that, the method that the piece ratio is omitted in the described prediction of step a is: determine the omission piece in the present encoding row, omit the number of piece in the statistics present encoding row, the number of the omission piece added up divided by the piece sum in the present encoding row, is obtained omitting the omission piece ratio of piece ratio as the next line piece.
4, coding method according to claim 1, it is characterized in that, in a frame the quantity of code line greater than 1 o'clock, the method that the piece ratio is omitted in the described prediction of step a is: determine before the present encoding row adjacent N omission piece capable, add up the omission piece number of described N in capable, with the number of the omission piece added up divided by described N capable in the sum of piece, obtain omitting the omission piece ratio of piece ratio as the next line piece, described N is less than or equal to the quantity of described code line and greater than 1 natural number.
5, according to claim 3 or 4 described coding methods, it is characterized in that, the method of described definite omission piece is: the motion vector that obtains piece, and obtain the compressed symbolic of this piece, if the motion vector of this piece that is obtained and compressed symbolic are 0, determine that then this piece is the omission piece, otherwise determine that this piece is not to omit piece.
6, coding method according to claim 1 is characterized in that, described estimated rate is 10%～40%.
7, coding method according to claim 6 is characterized in that, described estimated rate is 25%.
8, coding method according to claim 1 is characterized in that, described size is 16 * 16 pixels.
9, coding method according to claim 8 is characterized in that, described omission block type code word is 010; Described sub-block type is 16 * 16 pixels, and corresponding code word is 1; Described sub-block type is 8 * 8 pixels, and corresponding code word is 011; Described sub-block type is 16 * 8 pixels, and corresponding code word is 00100; Described sub-block type is 8 * 16 pixels, and corresponding code word is 00101.
10, coding method according to claim 1 is characterized in that, the described current piece that is encoded is not that further comprise: compressed symbolic and motion vector information to this piece are encoded when omitting piece.
11, coding method according to claim 1 is characterized in that, further comprises specifically comprising decode procedure:
Utilize the pairing decoding process of coded system of each frame video data first row being encoded that first row of each frame video data is decoded with described cataloged procedure;
Predict the omission piece ratio in the next line video data of current decoded row in this frame;
According to predicting the outcome of obtaining, judge whether the omission piece ratio in the next line of being predicted is less than or equal to estimated rate, if, then adopt in the described cataloged procedure to be less than or equal to code word identical under the estimated rate situation, omission piece and encoding block in the next line video data are decoded at described omission piece ratio; If not, then adopt the decoding process of the standard code mode correspondence described in the cataloged procedure that the next line video data is decoded.
According to claim 1 or 11 described coding methods, it is characterized in that 12, described standard code mode is standard skip_run coded system H.26L.
13, coding method according to claim 11 is characterized in that, the estimated rate in the described decode procedure equals the estimated rate in the described cataloged procedure.
14, coding method according to claim 11 is characterized in that, the piece that described current decoding obtains is not that further comprise: compressed symbolic and motion vector information to this piece are decoded when omitting piece.
15, coding method according to claim 1, it is characterized in that, described predefined coded system is: whether piece is to omit piece according to being encoded, utilize and omit piece as the pairing omission block type of codified block type code word, or the pairing code word of sub-block type of non-omission piece, piece is encoded to being encoded.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100053476A CN100399828C (en) | 2005-02-02 | 2005-02-02 | Video-frequency data encoding method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100053476A CN100399828C (en) | 2005-02-02 | 2005-02-02 | Video-frequency data encoding method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1816148A true CN1816148A (en) | 2006-08-09 |
| CN100399828C CN100399828C (en) | 2008-07-02 |
Family
ID=36908066
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2005100053476A Active CN100399828C (en) | 2005-02-02 | 2005-02-02 | Video-frequency data encoding method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100399828C (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102640495A (en) * | 2009-09-10 | 2012-08-15 | Sk电信有限公司 | Motion vector encoding/decoding method and device and image encoding/decoding method and device using same |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5552832A (en) * | 1994-10-26 | 1996-09-03 | Intel Corporation | Run-length encoding sequence for video signals |
| JP2000184203A (en) * | 1998-12-14 | 2000-06-30 | Ricoh Co Ltd | Information processor |
| US6859500B2 (en) * | 2001-03-20 | 2005-02-22 | Telefonaktiebolaget Lm Ericsson | Run-length coding of non-coded macroblocks |
| CN100452883C (en) * | 2001-12-17 | 2009-01-14 | 微软公司 | Skip macroblock coding |
| CN1332563C (en) * | 2003-12-31 | 2007-08-15 | 中国科学院计算技术研究所 | Coding method of video frequency image jump over macro block |
-
2005
- 2005-02-02 CN CNB2005100053476A patent/CN100399828C/en active Active
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102640495A (en) * | 2009-09-10 | 2012-08-15 | Sk电信有限公司 | Motion vector encoding/decoding method and device and image encoding/decoding method and device using same |
| CN102640495B (en) * | 2009-09-10 | 2016-04-06 | Sk电信有限公司 | The image coding/decoding method of motion vector coding/decoding method and device and use the method and device and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100399828C (en) | 2008-07-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100348051C (en) | An enhanced in-frame predictive mode coding method | |
| CN101031086A (en) | Video-information encoding method and video-information decoding method | |
| CN1968418A (en) | System and method for image data processing using hybrid type | |
| CN1592412A (en) | Adaptive variable-length coding methods for image data | |
| CN1589023A (en) | Coding and decoding method and device for multiple coded list lengthening based on context | |
| CN1913642A (en) | Video encoding apparatus and video encoding method | |
| CN1784015A (en) | Inage predicting encoding method in frame | |
| CN1949877A (en) | Adaptive quantization controller and method thereof | |
| CN1665303A (en) | Method of entropy coding of transformation coefficient in image/video coding | |
| CN1725863A (en) | The method of intra-prediction and the equipment that are used for video data | |
| CN1675932A (en) | System and method for rate-distortion optimized data partitioning for video coding using backward adapatation | |
| CN1757237A (en) | Video encoding | |
| CN1893649A (en) | Method and apparatus for operational frame-layer rate control in video encoder | |
| CN1527610A (en) | Entropy encoding method for encoding video predictive residual error coefficient | |
| CN1526240A (en) | Method and device for generating a scalable coded video signal from a non-scalable coded video signal | |
| CN101895756A (en) | Method and system for coding, decoding and reconstructing video image blocks | |
| CN1209928C (en) | Inframe coding frame coding method using inframe prediction based on prediction blockgroup | |
| CN1193621C (en) | Approximate IDCT for scalable video and image decoding of computational complexity | |
| CN1268136C (en) | Frame field adaptive coding method based on image slice structure | |
| CN100348053C (en) | Error masking method using moving picture electronic watermark of video | |
| CN1878308A (en) | Orthogonal transformation method for image and video compression | |
| CN1816148A (en) | Video-frequency data encoding method | |
| CN101043632A (en) | Image coding/decoding method and system | |
| CN1925616A (en) | Macro block pair class frame field adaptive coding/decoding method | |
| CN1224273C (en) | Video encoder and recording apparatus |
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 |