CN1564604A - Gradation tree set partitioning image compression method based on tree shaped structure - Google Patents

Gradation tree set partitioning image compression method based on tree shaped structure Download PDF

Info

Publication number
CN1564604A
CN1564604A CN 200410017558 CN200410017558A CN1564604A CN 1564604 A CN1564604 A CN 1564604A CN 200410017558 CN200410017558 CN 200410017558 CN 200410017558 A CN200410017558 A CN 200410017558A CN 1564604 A CN1564604 A CN 1564604A
Authority
CN
China
Prior art keywords
tree
stream
tree structure
threshold value
encoded
Prior art date
Application number
CN 200410017558
Other languages
Chinese (zh)
Inventor
华赟
胡波
徐晟�
高佳
Original Assignee
复旦大学
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 复旦大学 filed Critical 复旦大学
Priority to CN 200410017558 priority Critical patent/CN1564604A/en
Publication of CN1564604A publication Critical patent/CN1564604A/en

Links

Abstract

The invention includes following steps: the encode end first gets the distribution of image energy in time frequency domain through transform of discrete wavelet, the partition in hierarchical tree is set for each level of wavelet coefficient according to relevance between each wavelets coefficients; the wavelet coefficient of each tree is separately made SPIHT encode, the result of encode is temporarily stores in the encode end, finally the encoding result of each tree is combined into a data flow to use for storage and transmission. The decode process is the reverse process of encode.

Description

基于树状结构的等级树集合划分视频图像压缩方法 Video Compression set partitioning method based on a hierarchical tree of the tree structure

技术领域 FIELD

本发明属于视频图像压缩技术领域,具体涉及一种基于树状结构的等级树集合划分视频图像压缩方法。 The present invention belongs to the field of video image compression technology, particularly relates to a hierarchical tree based on the tree structure set partitioning method for video compression.

背景技术 Background technique

等级树集合划分(SPIHT)算法充分考虑了数据之间的相关性,并且在编码时还考虑了同一数据中高比特数据重要性高于低比特数据的特性。 Set Partitioning in Hierarchical Trees (SPIHT) algorithm full account of the correlation between the data and the same data considered of high importance greater than the lower bit data of the characteristic data coding bit Shihai. 所以使用SPIHT方法来压缩、解压缩视频图像可以得到比较高的压缩比而不增加解压缩结果的失真度,所以该方法受到了日益广泛的关注。 So using SPIHT method to compress, decompress video images can be relatively high compression ratio without increasing distortion decompression result, this method has been increasingly widespread attention. 在具体实现的过程中,编码系统需要建立三个链表,即重要像素链表(LSP)、不重要像素链表(LIP)和不重要像素集合链表(LIS),这三个链表用来记录树状结构分裂的中间数据。 In the specific implementation process, the encoding system needs to establish three lists, i.e., lists significant pixels (the LSP), the list is not important pixels (LIP) and insignificant pixel set list (the LIS), which tree structure is used to record lists three intermediate data division. 通过链表的使用,编码码流可以按照阈值(重要性)下降的顺序排列,从而保证重要信息的传输而可以截断非重要信息,得到任意截断码流的高压缩比的压缩效果。 By using the list of encoded code stream may be arranged in the order of a threshold value (importance) of the drop, thereby ensuring the transmission of important information and non-critical information may be truncated to obtain any truncated codestream effect high compression ratio. 为了提高压缩效果,要求树状结构包括更多的数据。 In order to improve the compression effect, comprising a tree structure requires more data. 但是随着树状结构中数据量的增加,三个链表的长度就越来越长,在实际应用中就要求有巨大的内存空间,这就增加了系统的成本和复杂度。 But with the increasing amount of data in a tree structure, the length of three of the list is getting longer, in practical applications will require a huge memory space, which increases the cost and complexity of the system. 所以在不降低压缩效果的前提下,缩短链表长度的方法正成为研究的热点。 Therefore, without lowering the compression effect, a method to shorten the chain length becoming a hot research.

发明内容 SUMMARY

本发明的目的是提出一种基于树状结构的等级树集合划分(SPIHT)视频图像压缩方法,以保证压缩效果不下降的前提下,大大缩短链表的长度,节省系统的内存空间开销。 Object of the present invention is to propose a hierarchical tree based on the set partitioning tree (SPIHT) video compression method, compression to ensure the premise does not drop, greatly reducing the length of the list, save memory space overhead of the system.

本发明提出的基于树状结构的等级树集合划分(SPIHT)视频图像压缩方法,编码的具体步骤如下:首先通过离散小波变换得到图像能量在时频率域上的分布,由于图像的平滑性,图像能量集中在低频部分;根据小波系数之间的相关性,将各级的小波系数按照树状结构进行划分;然后对每棵树的小波系数分别进行SPIHT编码,编码结果分别暂时存放在编码端;最后将每棵树的编码结果合成为一个码流用于存储或者传输。 Specific steps set partitioning (SPIHT) video compression method, based on a hierarchical tree coding tree structure proposed in the present invention are as follows: First, an image obtained when the energy distribution in the frequency domain by a discrete wavelet transform, since the smoothness of the image, the image energy is concentrated in the low frequency portion; according to the correlation between the wavelet coefficients, wavelet coefficients are divided according to levels of the tree structure; then, for each tree SPIHT wavelet coefficients are coded, the coding results are temporarily stored in an encoder; Finally, the result of coding each tree into one stream for transmission or storage.

根据小波系数之间的数据相关性,将各级的小波系数按照树状结构进行划分是指以最低频子带的每个系数为树根,按照不同级别之间小波系数位置的数据相关性得到树状结构中每个点的数据。 Correlation, wavelet coefficients at all levels divide each coefficient refers to the lowest frequency subband according to the root tree, according to the data position of the wavelet coefficients of correlation between the different levels between the data obtained wavelet coefficients tree structure data for each point. 树状结构中,上一级小波系数和下一级小波系数之间的关系称为父母和子女或后代的关系。 Tree structure, the relationship between the upper level and the next level of wavelet coefficients of wavelet coefficients referred to the relationship between parents and children or offspring. 在小波系数中,不同子带相同位置的系数,往往在数值上有相似性,根据这样的关系,将最低频子带的每个系数作为树的根节点,高一级的子代中相同位置的系数作为树状结构的第一级子女,更高一级的子代中与每个第一级的子女相同位置的系数作为第一级子女的子女,也是树状结构的第二级子女,……直到最高频子带的系数作为最后一级的子女。 In the wavelet coefficients, the coefficients of the same position of different sub-bands, often have similar properties numerically, according to such a relationship, the coefficients of each sub-band of the lowest frequency as the root node of the tree, one of offspring in the same position children coefficients as the first stage of the tree structure, higher level of progeny with the same coefficient for each position of the first stage of the first stage of the child as the child of the child, the child is the second stage of the tree structure, ...... until the coefficient of the highest frequency sub-bands as the last stage of their children.

对每棵树的小波系数分别进行SPIHT编码,可以减少同时处理的小波系数,产生的中间结果较少,缩短了重要像素链表(LSP)、不重要像素链表(LIP)和不重要像素集合链表(LIS)的长度。 For each tree SPIHT wavelet coefficients are coded, can be reduced simultaneously processed wavelet coefficients, fewer intermediate results produced a significant shortening pixel list (the LSP), the list is not important pixels (LIP) and insignificant pixel set list ( LIS) length. 其方法就是将每棵树的编码结果都按照阈值下降的顺序依次得到,直到阈值下降到可以满足压缩要求为止。 The method which is the encoding result of each tree are sequentially obtained in the order of descending threshold value, until the threshold down to meet the requirements up compression. 阈值下降极限可以由前一帧组的最小阈值或者经验阈值得到的预测阈值决定。 Threshold limit may be decreased by the minimum threshold or worth empirical threshold of the previous frame to the prediction group a threshold determination. 每棵树的小波系数进行SPIHT编码的结果不予直接传输,而是暂存在编码端,存放时将各阈值情况下的编码码流依次存放,并且记录各阈值情况下的编码码流长度。 The results of each tree SPIHT coding wavelet coefficients not transmitted directly, but is temporarily stored in the encoder, when storing the encoded bit stream at each threshold value are sequentially stored, and the recording length in the encoded bit stream for each threshold value.

在所有的树状结构的小波系数编码结束后,为了得到符合压缩比要求的目标码流,需要将每棵树的编码结果合成为目标码流。 At the end of the encoding of all wavelet coefficients of the tree structure, in line with the compression ratio in order to obtain required target stream, each tree needs to be synthesized as a target of the coding result stream. 合成码流的方法是确定最小的阈值,称为截断阈值,使得每棵树编码码流中不小于该阈值的码流之和不大于目标码流长度,将这些编码码流和码流长度合成为目标码流,剩余的目标码流再由每棵树的其余编码码流平均分配。 The method of synthesis of the code stream is determined minimum threshold value, referred to as a cut-off threshold, such that each tree encoded bit stream is not smaller than the threshold value of the code stream is not greater than the target length of the stream, the encoded code stream and the code stream length synthesis of these the target stream, the target remaining stream flows evenly again by the other encoded per tree. 就是将每棵树编码结果中阈值不小于截断阈值的码流和这些码流的长度直接作为目标码流,目标码流不足的部分由每棵树编码结果中阈值小于截断阈值的码流平均分配。 Is the length of the encoding result for each tree threshold value not less than the cutoff threshold of streams and these streams directly as an object code stream, stream cutoff threshold equal distribution portion is less than a target code stream is less than the coding result for each tree threshold value .

编码过程的重点在于树状结构的划分、树状结构小波系数编码结果的存放和目标码流的合成。 Key encoding process is to divide the tree structure, synthesis target storage and stream encoding wavelet coefficients in a tree structure results.

在解码端,解码过程是编码过程的逆过程:首先将待解码的码流分配给每棵树的缓存,再对每棵树分配到的码流依次进行SPIHT解码,得到树状结构的小波系数,再将树状结构的小波系数还原为按子带排布的小波系数,通过小波逆变换得到解码图像。 The decoding side, the decoding process is the reverse process of the encoding process: First, the code stream to be decoded is allocated to the buffer of each tree, and then assigned to each tree to decode the code stream sequentially SPIHT obtain wavelet coefficients in a tree structure , then the tree structure of wavelet coefficients of wavelet coefficients are reduced to sub-bands are arranged, the decoded image obtained by the inverse wavelet transform.

本发明所提出的基于树状结构的等级树集合划分(SPIHT)视频图像压缩方法,有效的解决了图像数据量和链表长度之间的矛盾。 Set partitioning (SPIHT) video compression based on a hierarchical tree of the tree structure of the proposed method of the present invention effectively solves the conflict between the chain length and the amount of image data. 为了提高压缩效果,可以将多帧的图像(帧组)一起进行离散小波变换,使得每棵树可以包括足够多的小波系数;由于每棵树分别编码,并不会导致重要像素链表(LSP)、不重要像素链表(LIP)和不重要像素集合链表(LIS)长度的过度加长。 In order to increase compression, the image (frame group) the plurality of frames may be performed together discrete wavelet transform, so that each tree may include a sufficient number of wavelet coefficients; as encode each tree, and does not cause significant list pixels (LSP) excessive chain extension unimportant pixels (LIP) and insignificant pixel set list (LIS) length.

具体实施方式 Detailed ways

一、编码端的具体实施方法1、离散小波变换结果的树状结构划分离散小波变换可以使用三维的离散小波变换,即在行方向、列方向和时间方向分别进行离散小波变换。 A specific embodiment of the method of the encoder 1, the tree structure into discrete wavelet transform may be a discrete wavelet transform result using a three dimensional discrete wavelet transform, i.e. in the row direction, the column direction and the time direction, discrete wavelet transform, respectively. 变换结果的最低频每个系数作为一棵树的根节点,并且按照下面的关系,构成树状结构。 The results of each lowest-frequency transform coefficients as the root of a tree, and in accordance with the following relationship, constituting the tree structure. 假设最低频系数的大小为Wmin×Hmin,其中Wmin和Hmin分别是最低频帧的最低频子带的宽度和高度。 Assuming that the size of the lowest frequency coefficient Wmin × Hmin, wherein Hmin and Wmin are the lowest frequency of the lowest frequency subband frame width and height.

1)根节点子女寻找方法 其子女为: 1) children of the root node is looking for ways to their children:

2)二维子女寻找方法 其子女为:{(2x,2y,z),(2x+1,2y,z),(2x,2y+1,z),(2x+1,2y+1,z)} (2)3)三维子女寻找方法 其子女为: 2) two-dimensional children as their children looking for ways: {(2x, 2y, z), (2x + 1,2y, z), (2x, 2y + 1, z), (2x + 1,2y + 1, z )} (2) 3) dimensional children as their children looking for ways:

2、每棵树状结构小波系数SPIHT编码结果的存放DM,N表示阈值从2N+1下降到2N时第M棵树阈值为N的编码数据。 2, each of them SPIHT encoded result of the tree structure of wavelet coefficients of the DM is stored, encoded data N represents the threshold value drops from 2N to 2N + 1 M-N of the threshold tree. LM,N表示阈值从2N+1下降到2N时第M棵树阈值为N的编码数据长度。 LM, N denotes the length of the coded data when the threshold value drops from 2N to 2N + 1 M-N of the threshold tree. 所有树的编码结果存放的格式如下: All format encoded result stored in the tree as follows:

3、目标码流的合成如果有M棵树,要求的目标码流长度为Q。 3, compositing target stream if M tree, the target length of the stream required is Q. 在阈值降到2P时,所有树的总码流长度为N1=&Sigma;n=1M&Sigma;m=PNLn,m,]]>在阈值降到2P-1时,所有树的总码流长度为N2=&Sigma;n=1M&Sigma;m=P-1NLn,m,]]>并且满足N1<Q≤N2,那么2P即为截断阈值。 When reducing the threshold to 2P, the total length of the stream all the trees is N1 = & Sigma; n = 1M & Sigma; m = PNLn, m,]]> threshold drops 2P-1, the total length of the stream all the trees is N2 = & Sigma; n = 1M & Sigma; m = P-1NLn, m,]]> and satisfies N1 <Q≤N2, then the 2P is the cutoff threshold. 先将M棵树阈值降到2P时的所有码流和码流长度作为目标码流,如果每棵树进入目标码流的码流长度要用X比特表示,此时目标码流约为N1+X,再将剩余的Q-N1-X的目标码流平均分配到M棵树中其余的编码码流中,也就是将每棵树阈值为2P-1的前(Q-N1-X)/M码流作为目标码流。 First of all code stream and the code stream length M of the tree down to the threshold value as the target code stream 2P, if each tree into the target code length of the stream bit stream X represents a use where the target stream is about N1 + X, then the remaining target code Q-N1-X of the flow evenly distributed in the rest of the tree M encoded bit stream, i.e. the threshold value of each tree before 2P-1 (Q-N1-X) / M stream stream as an object code. 具体在目标码流中,各棵树的编码码流是这样安排的:第1棵树&Sigma;m=PNL1,m,]]>D1,N——D1,P;第2棵树&Sigma;m=PNL2,m,]]>D2,N——D2,P;…… In certain specific bit stream, each of the encoded code stream is organized as a tree: a first tree & Sigma; m = PNL1, m,]]> D1, N - D1, P; a second tree & Sigma; m = PNL2, m,]]> D2, N - D2, P; ......

第M棵树&Sigma;m=PNLM,m,]]>DM,N——DM,P;D1,P-1中的第1个比特;D2,P-1中的第1个比特;……DM,P-1中的第1个比特;D1,P-1中的第2个比特;D2,P-1中的第2个比特;……DM,P-1中的第2个比特;……直到目标码流长度达到要求。 M-tree & Sigma; m = PNLM, m,]]> DM, N - DM, P; D1, P-1 in the first bit; D2, P-1 in the first bit; ...... DM, P-1 in the first bit; D1, P-1 in the second bit; D2, P-1 in the second bit; ...... DM, P-1 in the second bit; ...... until the target code stream length requirements.

二、解码端的具体实施方法解码的过程完全为编码的逆过程。 Second, the specific embodiment of a decoding method for decoding process ends completely reverse the encoding process. 首先将待解码的码流分配给每棵树的缓存,再对每棵树分配到的码流依次进行SPIHT解码,得到树状结构的小波系数,再将树状结构的小波系数还原为按子带排布的小波系数,通过小波逆变换得到解码图像。 First, the code stream is allocated to be decoded to the buffer of each tree, and then assigned to each tree to decode the code stream sequentially SPIHT obtain wavelet coefficients in a tree structure, the tree structure of wavelet coefficients and then reduced to the presser band wavelet coefficients arranged, the decoded image obtained by the inverse wavelet transform.

三、仿真的结果具体的仿真条件如下:Miss American视频图像组1-8帧图像的Y值数据,每帧图像大小为352×288。 Third, the simulation results of specific simulation conditions are as follows: Y value of the frame image data 1-8 Miss American video image group, the image size of each frame is 352 × 288. 进行三级三维离散小波变换,再对低频帧进行两级二维离散小波变换,小波基选用Daubechies9/7双正交小波(行方向和列方向)和Haar小波(时间方向)。 For three dimensional discrete wavelet transform, and then the low frame two dimensional discrete wavelet transform, Wavelet selected Daubechies9 / 7 biorthogonal wavelet (row direction and column direction) and the Haar wavelet (time direction). 共有99棵树。 A total of 99 trees.

实验结果如下: The results are as follows:

*优化后的LIP、LIP、LIS是99棵树中最大的长度,并且每棵树编码都进行到阈值降为8为止,试验证明阈值降到8,一般就能满足压缩比的要求。 * LIP optimized, LIP, LIS 99 is the maximum length of the tree, and each tree are coded to be reduced until the threshold value 8, the test proved that the threshold value below 8, generally able to meet the requirements of the compression ratio.

通过上面的实验结果我们可以发现,本SPIHT编码方法的结果虽然降低了PSNR(降低得非常小),但是用于存储链表的空间可以大大的减小。 We can see by the above experimental results, the results of this coding method SPIHT while lowering the PSNR (very small reduction), but the space for storing the linked list can be greatly reduced.

Claims (5)

1.一种基于树状结构的等级树集合划分视频图像压缩方法,其特征在于编码的具体步骤如下:首先通过离散小波变换得到图像能量在时频率域上的分布,根据小波系数之间的相关性,将各级的小波系数按照树状结构进行划分;然后对每棵树的小波系数分别进行SPIHT编码,编码结果分别暂时存放在编码端;最后将每棵树的编码结果合成为一个码流用于存储或者传输。 A tree structure based on a hierarchical tree set partitioning video compression method, wherein the step of encoding specific as follows: when the distribution is first obtained in the frequency domain by a discrete wavelet transform image energy, wavelet coefficients according to the correlation between of the wavelet coefficients in accordance with levels of the tree structure divided; and each tree SPIHT wavelet coefficients are coded, the coding results are temporarily stored in the encoding side; final encoding result of each tree into one stream with in stored or transmitted.
2.根据权利要求1所述的基于树状结构的等级树集合划分视频图像压缩方法,其特征在于所述树状结构的划分是以最低频子带的每个系数为树根,按照不同级别之间小波系数位置的数据相关性得到树状结构中每个点的数据。 2. Based on the set of a hierarchical tree of the tree structure according to claim 1 divided video compression method, wherein said tree structure dividing each coefficient is the lowest frequency subband is the root, according to their levels wavelet coefficients between the location of correlation data obtained for each point in the tree structure.
3.根据权利要求1所述的一种基于树状结构的等级树集合划分视频图像压缩方法,其特征在于每棵树都分别进行SPIHT编码,每棵树的编码结果都按照阈值下降的顺序得到,直到阈值下降到可以满足压缩要求为止。 According to one of the claim 1, a hierarchical tree based on the tree structure set partitioning method for video compression, wherein each tree SPIHT encoded separately, each tree results have been coded in order of descending threshold until the compression threshold down to meet the requirements so far.
4.根据权利要求1或3所述的一种基于树状结构的等级树集合划分视频图像压缩方法,其特征在于每棵树的编码结果先暂存在编码端,存放时将各阈值情况下的编码码流依次存放,并且记录各阈值情况下的编码码流长度。 According to one or more of the claims 1 to 3 divided set of video compression method based on a hierarchical tree of the tree structure, wherein each of the first tree of the encoding result is temporarily stored in the encoder, when storing the threshold value in each of the sequentially storing the encoded bit stream, and recording the encoded code stream length at each threshold value.
5.根据权利要求1或4所述的一种基于树状结构的等级树集合划分视频图像压缩方法,其特征在于合成码流的方法是确定最小的阈值,使得每棵树编码码流中不小于该阈值的码流之和不大于目标码流长度,将这些编码码流和码流长度合成作为目标码流,剩余的目标码流再由每棵树的其余编码码流平均分配。 According to one or more of claim 1 to claim 4 based on a hierarchical tree of the tree structure set partitioning video compression method, wherein the code stream is synthesized to determine the minimum threshold value, the encoded bit stream such that each tree is not smaller than the threshold value of the code stream is not greater than the target length of the stream, the stream and the encoded code stream length synthetic target stream, the target remaining stream flows evenly again by the other encoded per tree.
CN 200410017558 2004-04-08 2004-04-08 Gradation tree set partitioning image compression method based on tree shaped structure CN1564604A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 200410017558 CN1564604A (en) 2004-04-08 2004-04-08 Gradation tree set partitioning image compression method based on tree shaped structure

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 200410017558 CN1564604A (en) 2004-04-08 2004-04-08 Gradation tree set partitioning image compression method based on tree shaped structure

Publications (1)

Publication Number Publication Date
CN1564604A true CN1564604A (en) 2005-01-12

Family

ID=34479040

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 200410017558 CN1564604A (en) 2004-04-08 2004-04-08 Gradation tree set partitioning image compression method based on tree shaped structure

Country Status (1)

Country Link
CN (1) CN1564604A (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101682765B (en) 2007-05-30 2012-04-11 三叉微系统(远东)有限公司 Method of determining an image distribution for a light field data structure
CN102473315A (en) * 2009-08-20 2012-05-23 汤姆森特许公司 Method and apparatus for reusing tree structures to encode and decode binary sets
CN103024399A (en) * 2013-01-18 2013-04-03 北京航空航天大学 Wavelet transform based extreme-low bit-rate video compressing and coding method
CN102187656B (en) * 2008-10-27 2014-10-01 日本电信电话株式会社 Pixel prediction value generation procedure automatic generation method, image encoding method, image decoding method, devices using these methods, programs for these methods, and recording medium on which these programs are recorded
CN106612437A (en) * 2016-01-11 2017-05-03 四川用联信息技术有限公司 Graphic image compression method

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101682765B (en) 2007-05-30 2012-04-11 三叉微系统(远东)有限公司 Method of determining an image distribution for a light field data structure
CN102187656B (en) * 2008-10-27 2014-10-01 日本电信电话株式会社 Pixel prediction value generation procedure automatic generation method, image encoding method, image decoding method, devices using these methods, programs for these methods, and recording medium on which these programs are recorded
US9124289B2 (en) 2008-10-27 2015-09-01 Nippon Telegraph And Telephone Corporation Predicted pixel value generation procedure automatic producing method, image encoding method, image decoding method, apparatus therefor, programs therefor, and storage media which store the programs
CN102473315A (en) * 2009-08-20 2012-05-23 汤姆森特许公司 Method and apparatus for reusing tree structures to encode and decode binary sets
CN103024399A (en) * 2013-01-18 2013-04-03 北京航空航天大学 Wavelet transform based extreme-low bit-rate video compressing and coding method
CN106612437A (en) * 2016-01-11 2017-05-03 四川用联信息技术有限公司 Graphic image compression method

Similar Documents

Publication Publication Date Title
RU2607249C2 (en) Method and apparatus for encoding residual block, method and apparatus for decoding residual block
US6546143B1 (en) Efficient wavelet-based compression of large images
CN101133649B (en) Encoding device, encoding method, decoding device and decoding method
US5881176A (en) Compression and decompression with wavelet style and binary style including quantization by device-dependent parser
KR101375668B1 (en) Method and apparatus for encoding transformed coefficients and method and apparatus for decoding transformed coefficients
CN1135707C (en) Method and apparatus for compression using reversible sub-wave conversion and embedded code stream
JP2018067946A (en) Methods and apparatus for video encoding and video decoding binary sets using adaptive tree selection
CN102316320B (en) Skip macroblock coding
Sudhakar et al. Image compression using coding of wavelet coefficients–A survey
CN1299243C (en) Image coding and decoding method, image coding and decoding device
JP4208900B2 (en) Decoding method and decoding apparatus
US7440626B2 (en) Image transcoding
KR100527690B1 (en) Signal encoding method and apparatus and decoding method and apparatus
JP3204645B2 (en) Video encoding and decoding method and apparatus
US7359561B2 (en) Picture encoding with wavelet transform and block area weights
CN1266649C (en) Video coding method
JP2005304056A (en) Method and device of encoding
AU719715B2 (en) Data compression using adaptive bit allocation and hybrid lossless entropy encoding
CN102474614B (en) Video encoding method and apparatus and video decoding method and apparatus, based on hierarchical coded block pattern information
Askelöf et al. Region of interest coding in JPEG 2000
JP3853758B2 (en) Image encoding device
JP4273996B2 (en) Image encoding apparatus and method, and image decoding apparatus and method
JPWO2008143156A1 (en) Encoding apparatus, encoding method, decoding apparatus, and decoding method
JP3800551B2 (en) Data processing apparatus and method
CN104581163B (en) Method and apparatus for encoding and decoding image by using large transformation unit

Legal Events

Date Code Title Description
C06 Publication
C02 Deemed withdrawal of patent application after publication (patent law 2001)