This paper presents a simple and effective post-processing method for compressed images. This work focuses on the cyclic time-variance introduced by block-based and subband transform coders. We propose an algorithm to (almost) restore stationarity to the cyclo-stationary output of the conventional transform coders. Despite a simple, non-iterative structure, this method outperforms other methods of image enhancement known to us, e.g. linear and nonlinear filtering, projection on convex sets (POCS), wavelet-based and optimization-based methods. In particular, the proposed method performs very well in suppressing both blocking and ringing artifacts. Furthermore, it admits a solution with successive approximation. The resulting embeddedness is very useful for multimedia applications such as image browsing on the World Wide Web.
{"title":"Embedded post-processing for enhancement of compressed images","authors":"Aria Nosratinia","doi":"10.1109/DCC.1999.755655","DOIUrl":"https://doi.org/10.1109/DCC.1999.755655","url":null,"abstract":"This paper presents a simple and effective post-processing method for compressed images. This work focuses on the cyclic time-variance introduced by block-based and subband transform coders. We propose an algorithm to (almost) restore stationarity to the cyclo-stationary output of the conventional transform coders. Despite a simple, non-iterative structure, this method outperforms other methods of image enhancement known to us, e.g. linear and nonlinear filtering, projection on convex sets (POCS), wavelet-based and optimization-based methods. In particular, the proposed method performs very well in suppressing both blocking and ringing artifacts. Furthermore, it admits a solution with successive approximation. The resulting embeddedness is very useful for multimedia applications such as image browsing on the World Wide Web.","PeriodicalId":103598,"journal":{"name":"Proceedings DCC'99 Data Compression Conference (Cat. No. PR00096)","volume":"298 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132627938","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We improve upon previous results on the Burrows and Wheeler (BW)-algorithm. Based on the context tree model, we consider the specific statistical properties of the data at the output of the BWT. We describe six important properties, three of which have not been described elsewhere. These considerations lead to modifications of the coding method, which in turn improve the coding efficiency. We briefly describe how to compute the BWT with low complexity in time and space, using suffix trees in two different representations. Finally, we present experimental results about the compression rate and running time of our method, and compare these results to previous achievements.
我们改进了Burrows and Wheeler (BW)算法的先前结果。基于上下文树模型，我们考虑了BWT输出时数据的特定统计属性。我们描述了六个重要的性质，其中三个没有在其他地方描述。这些考虑导致了对编码方法的修改，从而提高了编码效率。我们简要描述了如何使用后缀树在两种不同的表示中计算低时间和空间复杂度的BWT。最后，给出了该方法的压缩率和运行时间的实验结果，并与前人的成果进行了比较。
{"title":"Modifications of the Burrows and Wheeler data compression algorithm","authors":"B. Balkenhol, S. Kurtz, Y. Shtarkov","doi":"10.1109/DCC.1999.755668","DOIUrl":"https://doi.org/10.1109/DCC.1999.755668","url":null,"abstract":"We improve upon previous results on the Burrows and Wheeler (BW)-algorithm. Based on the context tree model, we consider the specific statistical properties of the data at the output of the BWT. We describe six important properties, three of which have not been described elsewhere. These considerations lead to modifications of the coding method, which in turn improve the coding efficiency. We briefly describe how to compute the BWT with low complexity in time and space, using suffix trees in two different representations. Finally, we present experimental results about the compression rate and running time of our method, and compare these results to previous achievements.","PeriodicalId":103598,"journal":{"name":"Proceedings DCC'99 Data Compression Conference (Cat. No. PR00096)","volume":"1048 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131797779","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Achievable distortion bounds are derived for the cascade of structured families of binary linear channel codes and binary lattice vector quantizers. It is known that for the cascade of asymptotically good channel codes and asymptotically good vector quantizers the end-to-end distortion decays to zero exponentially fast as a function of the overall transmission rate, and is achieved by choosing a channel code rate that is independent of the overall transmission rate. We show that for certain families of practical channel codes and binary lattice vector quantizers, the overall distortion can still be made to decay to zero exponentially fast as the transmission rate grows, although the exponent is a sub-linear function of the transmission rate. This is achieved by carefully choosing a channel code rate that decays to zero as the transmission rate grows. Explicit channel code rate schedules are obtained for several well-known families of channel codes.
{"title":"Performance of quantizers on noisy channels using structured families of codes","authors":"A. Méhes, K. Zeger","doi":"10.1109/DCC.1999.755697","DOIUrl":"https://doi.org/10.1109/DCC.1999.755697","url":null,"abstract":"Achievable distortion bounds are derived for the cascade of structured families of binary linear channel codes and binary lattice vector quantizers. It is known that for the cascade of asymptotically good channel codes and asymptotically good vector quantizers the end-to-end distortion decays to zero exponentially fast as a function of the overall transmission rate, and is achieved by choosing a channel code rate that is independent of the overall transmission rate. We show that for certain families of practical channel codes and binary lattice vector quantizers, the overall distortion can still be made to decay to zero exponentially fast as the transmission rate grows, although the exponent is a sub-linear function of the transmission rate. This is achieved by carefully choosing a channel code rate that decays to zero as the transmission rate grows. Explicit channel code rate schedules are obtained for several well-known families of channel codes.","PeriodicalId":103598,"journal":{"name":"Proceedings DCC'99 Data Compression Conference (Cat. No. PR00096)","volume":"38 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114059027","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Life is based on two polymers, DNA and protein, whose properties can be described in a simple text file. It is natural to expect that standard text compression techniques would work on biological sequences as they do on English text. But biological sequences have a fundamentally different structure from linguistic ones, and standard compression schemes exhibit disappointing performance on them. We describe a new approach to compression that takes account of the underlying biochemical principles. This gives rise to a generalization of blending for statistical compressors where every context is used, weighted by its similarity to the current context. Results support what research in bioinformatics has shown, that there is little Markov dependency in protein. This cripples data compression schemes and reduces them to order zero models.
{"title":"Protein is incompressible","authors":"C. Nevill-Manning, I. Witten","doi":"10.1109/DCC.1999.755675","DOIUrl":"https://doi.org/10.1109/DCC.1999.755675","url":null,"abstract":"Life is based on two polymers, DNA and protein, whose properties can be described in a simple text file. It is natural to expect that standard text compression techniques would work on biological sequences as they do on English text. But biological sequences have a fundamentally different structure from linguistic ones, and standard compression schemes exhibit disappointing performance on them. We describe a new approach to compression that takes account of the underlying biochemical principles. This gives rise to a generalization of blending for statistical compressors where every context is used, weighted by its similarity to the current context. Results support what research in bioinformatics has shown, that there is little Markov dependency in protein. This cripples data compression schemes and reduces them to order zero models.","PeriodicalId":103598,"journal":{"name":"Proceedings DCC'99 Data Compression Conference (Cat. No. PR00096)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114190945","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We address the problem of distributed source coding, i.e. compression of correlated sources that are not co-located and/or cannot communicate with each other to minimize their joint description cost. In this work we tackle the related problem of compressing a source that is correlated with another source which is available only at the decoder. In contrast to prior information-theoretic approaches, we introduce a new construction and practical framework for tackling the problem based on the judicious incorporation of channel coding principles into this source coding problem. We dub our approach as distributed source coding using syndromes (DISCUS). We focus in this paper on trellis-structured constructions of the framework to illustrate its utility. Simulation results confirm the power of DISCUS, opening up a new and exciting constructive playing-ground for the distributed source coding problem. For the distributed coding of correlated i.i.d. Gaussian sources that are noisy versions of each other with "correlation-SNR" in the range of 12 to 20 dB, the DISCUS method attains gains of 7-15 dB in SNR over the Shannon-bound using "naive" independent coding of the sources.
{"title":"Distributed source coding using syndromes (DISCUS): design and construction","authors":"S. Pradhan, K. Ramchandran","doi":"10.1109/DCC.1999.755665","DOIUrl":"https://doi.org/10.1109/DCC.1999.755665","url":null,"abstract":"We address the problem of distributed source coding, i.e. compression of correlated sources that are not co-located and/or cannot communicate with each other to minimize their joint description cost. In this work we tackle the related problem of compressing a source that is correlated with another source which is available only at the decoder. In contrast to prior information-theoretic approaches, we introduce a new construction and practical framework for tackling the problem based on the judicious incorporation of channel coding principles into this source coding problem. We dub our approach as distributed source coding using syndromes (DISCUS). We focus in this paper on trellis-structured constructions of the framework to illustrate its utility. Simulation results confirm the power of DISCUS, opening up a new and exciting constructive playing-ground for the distributed source coding problem. For the distributed coding of correlated i.i.d. Gaussian sources that are noisy versions of each other with \"correlation-SNR\" in the range of 12 to 20 dB, the DISCUS method attains gains of 7-15 dB in SNR over the Shannon-bound using \"naive\" independent coding of the sources.","PeriodicalId":103598,"journal":{"name":"Proceedings DCC'99 Data Compression Conference (Cat. No. PR00096)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114374689","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
If x is a string of finite length over a finite alphabet A, let LZ(x) denote the length of the binary codeword assigned to x by the 1978 version of the Lempel-Ziv data compression algorithm, let t(x) be the number of phrases in the Lempel-Ziv parsing of x, and let /spl mu/(x) be the probability assigned to x by a memoryless source model. Using a very simple technique, we probe the pointwise redundancy bound LZ(x)+log/sub 2//spl mu/(x)/spl les/8t(x)max{-log/sub 2//spl mu/(a):a/spl isin/A}.
{"title":"A simple technique for bounding the pointwise redundancy of the 1978 Lempel-Ziv algorithm","authors":"J. Kieffer, E. Yang","doi":"10.1109/DCC.1999.755693","DOIUrl":"https://doi.org/10.1109/DCC.1999.755693","url":null,"abstract":"If x is a string of finite length over a finite alphabet A, let LZ(x) denote the length of the binary codeword assigned to x by the 1978 version of the Lempel-Ziv data compression algorithm, let t(x) be the number of phrases in the Lempel-Ziv parsing of x, and let /spl mu/(x) be the probability assigned to x by a memoryless source model. Using a very simple technique, we probe the pointwise redundancy bound LZ(x)+log/sub 2//spl mu/(x)/spl les/8t(x)max{-log/sub 2//spl mu/(a):a/spl isin/A}.","PeriodicalId":103598,"journal":{"name":"Proceedings DCC'99 Data Compression Conference (Cat. No. PR00096)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117316459","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
[Summary form only given]. This paper describes a fast, low-complexity, entropy efficient coder for wavelet pyramids. This coder approaches the entropy-limited coding rate of video wavelet pyramids, is fast in both hardware and software implementations, and has low complexity for use in ASICs. It consists of a modified Z-coder used to code the zero/non-zero significance function without adaptation. The wavelet pyramid is further sharpened by scaling to match the characteristics of the human visual system (HVS). We derive the statistical characteristics of quantized wavelet pyramids from NTSC video viewed under standard conditions. These video pyramids have substantial runs of zeros and also substantial runs of non-zeros. To explore these we developed a modification of the Z-coder and explored an application of it to code zero versus non-zero. Z-codecs have the advantage of a simple (no multipliers) and fast implementation combined with coding performance approximating that of an arithmetic codec. Our experiments showed that this coder compares favorably with straight arithmetic coding. Our encoder has significant speed advantage due to low cost implementation.
{"title":"Fast, modified Z-coding of wavelet pyramids","authors":"W. Lynch, K. Kolarov, Bill Arrighi","doi":"10.1109/DCC.1999.785695","DOIUrl":"https://doi.org/10.1109/DCC.1999.785695","url":null,"abstract":"[Summary form only given]. This paper describes a fast, low-complexity, entropy efficient coder for wavelet pyramids. This coder approaches the entropy-limited coding rate of video wavelet pyramids, is fast in both hardware and software implementations, and has low complexity for use in ASICs. It consists of a modified Z-coder used to code the zero/non-zero significance function without adaptation. The wavelet pyramid is further sharpened by scaling to match the characteristics of the human visual system (HVS). We derive the statistical characteristics of quantized wavelet pyramids from NTSC video viewed under standard conditions. These video pyramids have substantial runs of zeros and also substantial runs of non-zeros. To explore these we developed a modification of the Z-coder and explored an application of it to code zero versus non-zero. Z-codecs have the advantage of a simple (no multipliers) and fast implementation combined with coding performance approximating that of an arithmetic codec. Our experiments showed that this coder compares favorably with straight arithmetic coding. Our encoder has significant speed advantage due to low cost implementation.","PeriodicalId":103598,"journal":{"name":"Proceedings DCC'99 Data Compression Conference (Cat. No. PR00096)","volume":"417 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116181774","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. One of the most popular encoders in the literature is the LZ78, which was proposed by Ziv and Lempel (1978). We establish a recursive way to find the longest m-tuple match. We prove the following theorem that shows how to obtain a longest (m+1)-tuple match from the longest m-tuple match. It shows that a (m+1)-tuple match is the concatenation of the first (m-1) words of the m-tuple match with the next longest double match. Therefore, the longest (m+1)-tuple match can be found using the m-tuple match and a procedure to compute the longest double match. Our theorem is as follows. Let A be a source alphabet, let A* be the set of all finite strings of A, and D/spl sub/A*, such that if x/spl isin/D then all prefixes of x belong to D. Let u denote a one-sided infinite sequence. If b/sub 1//sup m/ is the longest m-tuple match in u, with respect to D, then there is a longest (m+1)-tuple match b/spl circ//sub 1//sup m+1/, such that b/spl circ//sub i/=b/sub i/,/spl forall/i/spl isin/{1,...m-1}. We implemented the fast mmLZ and the results show a improvement in compression of around 5% over the LZW, in the Canterbury Corpus (Arnold and Bell, 1997) with little extra computational cost.
{"title":"Fast multi-match Lempel-Ziv","authors":"M. Pinho, W. Finamore, W. Pearlman","doi":"10.1109/DCC.1999.785702","DOIUrl":"https://doi.org/10.1109/DCC.1999.785702","url":null,"abstract":"Summary form only given. One of the most popular encoders in the literature is the LZ78, which was proposed by Ziv and Lempel (1978). We establish a recursive way to find the longest m-tuple match. We prove the following theorem that shows how to obtain a longest (m+1)-tuple match from the longest m-tuple match. It shows that a (m+1)-tuple match is the concatenation of the first (m-1) words of the m-tuple match with the next longest double match. Therefore, the longest (m+1)-tuple match can be found using the m-tuple match and a procedure to compute the longest double match. Our theorem is as follows. Let A be a source alphabet, let A* be the set of all finite strings of A, and D/spl sub/A*, such that if x/spl isin/D then all prefixes of x belong to D. Let u denote a one-sided infinite sequence. If b/sub 1//sup m/ is the longest m-tuple match in u, with respect to D, then there is a longest (m+1)-tuple match b/spl circ//sub 1//sup m+1/, such that b/spl circ//sub i/=b/sub i/,/spl forall/i/spl isin/{1,...m-1}. We implemented the fast mmLZ and the results show a improvement in compression of around 5% over the LZW, in the Canterbury Corpus (Arnold and Bell, 1997) with little extra computational cost.","PeriodicalId":103598,"journal":{"name":"Proceedings DCC'99 Data Compression Conference (Cat. No. PR00096)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123920298","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. Current adaptive compression schemes such as GZIP and COMPRESS are impractical for database compression as they do not allow random access to individual records. A compression algorithm for general-purpose database systems must address the problem of randomly accessing and individually decompressing records, while maintaining compact storage of data. The SEQUITUR algorithm of Nevill-Manning et al., (1994, 1996, 1997) also adaptively compresses data, achieving excellent compression but with significant main-memory requirements. A preliminary version of SEQUITUR used a semi-static modelling approach to achieve slightly worse compression than the adaptive approach. We describe a new variant of the semi-static SEQUITUR algorithm, RAY, that reduces main-memory use and allows random-access to databases. RAY models repetition in sequences by progressively constructing a hierarchical grammar with multiple passes through the data. The multiple pass approach of RAY uses statistics on character pair repetition, or digram frequency, to create rules in the grammar. While our preliminary implementation is not especially fast, the multi-pass approach permits reductions in compression time, at the cost of affecting compression performance, by limiting the number of passes. We have found that RAY has practicable main-memory requirements and achieves better compression than an efficient Huffmann scheme and popular adaptive compression techniques. Moreover, our scheme allows random access to data and is not restricted to databases of text.
{"title":"A general-purpose compression scheme for databases","authors":"A. Cannane, H. Williams, J. Zobel","doi":"10.1109/DCC.1999.785676","DOIUrl":"https://doi.org/10.1109/DCC.1999.785676","url":null,"abstract":"Summary form only given. Current adaptive compression schemes such as GZIP and COMPRESS are impractical for database compression as they do not allow random access to individual records. A compression algorithm for general-purpose database systems must address the problem of randomly accessing and individually decompressing records, while maintaining compact storage of data. The SEQUITUR algorithm of Nevill-Manning et al., (1994, 1996, 1997) also adaptively compresses data, achieving excellent compression but with significant main-memory requirements. A preliminary version of SEQUITUR used a semi-static modelling approach to achieve slightly worse compression than the adaptive approach. We describe a new variant of the semi-static SEQUITUR algorithm, RAY, that reduces main-memory use and allows random-access to databases. RAY models repetition in sequences by progressively constructing a hierarchical grammar with multiple passes through the data. The multiple pass approach of RAY uses statistics on character pair repetition, or digram frequency, to create rules in the grammar. While our preliminary implementation is not especially fast, the multi-pass approach permits reductions in compression time, at the cost of affecting compression performance, by limiting the number of passes. We have found that RAY has practicable main-memory requirements and achieves better compression than an efficient Huffmann scheme and popular adaptive compression techniques. Moreover, our scheme allows random access to data and is not restricted to databases of text.","PeriodicalId":103598,"journal":{"name":"Proceedings DCC'99 Data Compression Conference (Cat. No. PR00096)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129286146","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. Constantinescu and Storer (1994) introduced an adaptive vector quantization algorithm (AVQ) that combines adaptive dictionary techniques with vector quantization (VQ). The algorithm typically equals or exceeds the compression of the JPEG standard on different classes of images and it often outperforms traditional trained VQ. We show how it is possible to improve AVQ on the class of images on which JPEG does best (i.e., "magazine photographs"). The improvement is possible by exploring the similarities in the dictionary built by AVQ. This is achieved by transforming the input vectors in a way similar to the one used in mean-shape-gain VQ (Oehler and Gray, 1993). In MSGVQ each vector x~/spl isin/R/sup n/ is decomposed as x~=g/spl middot/s~+E/sub x//spl middot/1~, where g=/spl par/x~-E/sub x//spl middot/1~/spl par/ and s~=(x~-E/sub x//spl middot/1~)/g; mean, gain and shape are quantized separately. We apply this idea to AVQ, changing the match heuristic: let and be respectively the of the dictionary block b and of the one anchored in p. The entry b is the best match if d(x~/sub p/,x/spl circ/)/spl les/T (x/spl circ/=g/sub p//spl middot/s~/sub b/+E/sub p//spl middot/1~) and its size is maximum. The triple is entropy coded and sent to the decoder. This simple modification of the match heuristic allows AVQ to improve the compression ratio on many images. In some cases this improvement is as high as 60%. Along with the better compression results, there is also an improvement in the overall visual quality of the decoded image, especially at high compression rate.
{"title":"Experiments with single-pass adaptive vector quantization","authors":"F. Rizzo, J. Storer, B. Carpentieri","doi":"10.1109/DCC.1999.785703","DOIUrl":"https://doi.org/10.1109/DCC.1999.785703","url":null,"abstract":"Summary form only given. Constantinescu and Storer (1994) introduced an adaptive vector quantization algorithm (AVQ) that combines adaptive dictionary techniques with vector quantization (VQ). The algorithm typically equals or exceeds the compression of the JPEG standard on different classes of images and it often outperforms traditional trained VQ. We show how it is possible to improve AVQ on the class of images on which JPEG does best (i.e., \"magazine photographs\"). The improvement is possible by exploring the similarities in the dictionary built by AVQ. This is achieved by transforming the input vectors in a way similar to the one used in mean-shape-gain VQ (Oehler and Gray, 1993). In MSGVQ each vector x~/spl isin/R/sup n/ is decomposed as x~=g/spl middot/s~+E/sub x//spl middot/1~, where g=/spl par/x~-E/sub x//spl middot/1~/spl par/ and s~=(x~-E/sub x//spl middot/1~)/g; mean, gain and shape are quantized separately. We apply this idea to AVQ, changing the match heuristic: let and be respectively the of the dictionary block b and of the one anchored in p. The entry b is the best match if d(x~/sub p/,x/spl circ/)/spl les/T (x/spl circ/=g/sub p//spl middot/s~/sub b/+E/sub p//spl middot/1~) and its size is maximum. The triple is entropy coded and sent to the decoder. This simple modification of the match heuristic allows AVQ to improve the compression ratio on many images. In some cases this improvement is as high as 60%. Along with the better compression results, there is also an improvement in the overall visual quality of the decoded image, especially at high compression rate.","PeriodicalId":103598,"journal":{"name":"Proceedings DCC'99 Data Compression Conference (Cat. No. PR00096)","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122731764","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：