Compression for Multimedia

by Irina Bocharova

Overview

Providing a thorough theoretical understanding of lossy compression techniques for image, video, speech, and audio compression, this book also covers the key features of each system, as well as practical applications, implementation issues, and design trade-offs. It presents comparisons of multimedia standards in terms of achieving known theoretical limits, whilst common and distinguishing features of the existing standards are explained and related to the background theory. There is detailed coverage of such topics as the H.264 video coding standard, low-complexity code-based vector quantizers, and the Blahut rate-distortion algorithm. Examples based on real multimedia data are also included, plus end-of-chapter problems to test understanding, algorithms that allow the reader to represent speech and audio signals efficiently, and an appendix on the basics of lossless coding. With an excellent balance of theory and practice, this book is ideal for undergraduate and graduate students, and is also a useful reference for practitioners.

Product Details

ISBN-13:	9780521114325
Publisher:	Cambridge University Press
Publication date:	12/24/2009
Pages:	280
Product dimensions:	6.80(w) x 9.70(h) x 0.80(d)

About the Author

Irina Bocharova is an Associate Professor at the Saint-Petersburg State University of Information Technologies, Mechanics and Optics. She has published over 50 technical papers and is the co-inventor of seven US patents in speech, video and audio coding. Her current research interests include convolutional codes, communication systems, source coding and its applications to speech, audio and image coding.

Table of Contents

Preface ix

1 Introduction 1

2 Analog to digital conversion 5

2.1 Analog and digital signals 5

2.2 Scalar quantization 15

2.3 Vector quantization 23

3 Elements of rate-distortion theory 42

3.1 Rate-distortion function 42

3.2 The Blahut algorithm 49

3.3 The Shannon lower bound and high-resolution approximations 54

3.4 Comparison of quantization procedures 58

3.5 Characteristics of digital speech, audio, image, and video signals 63

4 Scalar quantization with memory 66

4.1 Discrete-time filters 66

4.2 Linear predictive coding 78

5 Transform coding 91

5.1 Properties of transforms 92

5.2 The Karhunen-Loeve transform 95

5.3 The discrete Fourier transform 98

5.4 The discrete cosine transform 104

6 Filter banks and wavelet filtering 110

6.1 Linear filtering as a linear transform 110

6.2 "Moving average" filtering as a linear transform 112

6.3 "Moving difference" filtering as a linear transform 114

6.4 The Haar filter bank 115

6.5 Wavelet transform and wavelet filter bank 121

6.6 Hierarchical wavelet filtering as a linear transform. Properties of wavelet filtering 126

6.7 Historical background 132

6.8 Application of wavelet filtering to image compression 132

6.9 Embedded zerotree coding and set partitioning in hierarchical trees 135

7 Speech coding: techniques and standards 141

7.1 Direct sample-by-sample quantization: Standard G.711 145

7.2 ADPCM coders: Standards G.726, G.727 147

7.3 Linear prediction analysis-by-synthesis coders (LPC-AS) 152

7.4 LPC vocoders. MELP standard 162

7.5 An algorithm for computing linear spectral parameters 165

8 Image coding standards 171

8.1 Coding of bi-level fax images: JBIG Standard 172

8.2 Coding of halftone images: JPEG Standard 177

8.3 JPEG-LS 186

8.4 Standard JPEG-2000 188

9 Video-coding standards 197

9.1 Motion compensation method 197

9.2 Overview of video-coding standards 200

10 Audio-coding standards 223

10.1 Basics of perceptual coding 223

10.2 Overview of audio standards 227

A Lossless-coding techniques 238

A.1 Symbol-by-symbol lossless coding 238

A.2 Lossless block coding 241

A.3 Arithmetic coding 243

A.4 Decoding issues 246

A.5 Context lossless coding for sources with memory 249

A.6 QM-coder 251

A.7 Monotonic codes 254

A.8 Binarization 257

A.9 Adaptive coding 258

References 261

Index 265