Still Image and Video Compression with MATLAB E-Book

K. S. Thyagarajan

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Serie: Wiley - IEEE
• Sprache: Englisch

Seitenzahl: 523

Veröffentlichungsjahr: 2011

Beschreibung

This book describes the principles of image and video compression techniques and introduces current and popular compression standards, such as the MPEG series. Derivations of relevant compression algorithms are developed in an easy-to-follow fashion. Numerous examples are provided in each chapter to illustrate the concepts.

Leseprobe

Contents

COVER

HALF TITLE PAGE

TITLE PAGE

COPYRIGHT PAGE

DEDICATION

PREFACE

CHAPTER 1: INTRODUCTION

1.1 WHAT IS SOURCE CODING?

1.2 WHY IS COMPRESSION NECESSARY?

1.3 IMAGE AND VIDEO COMPRESSION TECHNIQUES

1.4 VIDEO COMPRESSION STANDARDS

1.5 ORGANIZATION OF THE BOOK

1.6 SUMMARY

REFERENCES

CHAPTER 2: IMAGE ACQUISITION

2.1 INTRODUCTION

2.2 SAMPLING A CONTINUOUS IMAGE

2.3 IMAGE QUANTIZATION

2.4 COLOR IMAGE REPRESENTATION

2.5 SUMMARY

REFERENCES

PROBLEMS

CHAPTER 3: IMAGE TRANSFORM

3.1 INTRODUCTION

3.2 UNITARY TRANSFORMS

3.3 KARHUNEN–LOÈVE TRANSFORM

3.4 PROPERTIES OF UNITARY TRANSFORMS

3.5 SUMMARY

REFERENCES

PROBLEMS

CHAPTER 4: DISCRETE WAVELET TRANSFORM

4.1 INTRODUCTION

4.2 CONTINUOUS WAVELET TRANSFORM

4.3 WAVELET SERIES

4.4 DISCRETE WAVELET TRANSFORM

4.5 EFFICIENT IMPLEMENTATION OF 1D DWT

4.6 SCALING AND WAVELET FILTERS

4.7 TWO-DIMENSIONAL DWT

4.8 ENERGY COMPACTION PROPERTY

4.9 INTEGER OR REVERSIBLE WAVELET

4.10 SUMMARY

REFERENCES

PROBLEMS

CHAPTER 5: LOSSLESS CODING

5.1 INTRODUCTION

5.2 INFORMATION THEORY

5.3 HUFFMAN CODING

5.4 ARITHMETIC CODING

5.5 GOLOMB–RICE CODING

5.6 RUN-LENGTH CODING

5.7 SUMMARY

REFERENCES

PROBLEMS

CHAPTER 6: PREDICTIVE CODING

6.1 INTRODUCTION

6.2 DESIGN OF A DPCM

6.3 ADAPTIVE DPCM

6.4 SUMMARY

REFERENCES

PROBLEMS

CHAPTER 7: IMAGE COMPRESSION IN THE TRANSFORM DOMAIN

7.1 INTRODUCTION

7.2 BASIC IDEA BEHIND TRANSFORM CODING

7.3 CODING GAIN OF A TRANSFORM CODER

7.4 JPEG COMPRESSION

7.5 COMPRESSION OF COLOR IMAGES

7.6 BLOCKING ARTIFACT

7.7 VARIABLE BLOCK SIZE DCT CODING

7.8 SUMMARY

REFERENCES

PROBLEMS

CHAPTER 8: IMAGE COMPRESSION IN THE WAVELET DOMAIN

8.1 INTRODUCTION

8.2 DESIGN OF A DWT CODER

8.3 ZERO-TREE CODING

8.4 JPEG2000

8.5 DIGITAL CINEMA

8.6 SUMMARY

REFERENCES

PROBLEMS

CHAPTER 9: BASICS OF VIDEO COMPRESSION

9.1 INTRODUCTION

9.2 VIDEO CODING

9.3 STEREO IMAGE COMPRESSION

9.4 SUMMARY

REFERENCES

PROBLEMS

CHAPTER 10: VIDEO COMPRESSION STANDARDS

10.1 INTRODUCTION

10.2 MPEG-1 AND MPEG-2 STANDARDS

10.3 MPEG-4

10.4 H.264

10.5 SUMMARY

REFERENCES

PROBLEMS

INDEX

STILL IMAGE AND VIDEO COMPRESSION WITH MATLAB

Copyright © 2011 by John Wiley & Sons, Inc. All rights reserved.

Published by John Wiley & Sons, Inc., Hoboken, New JerseyPublished simultaneously in Canada

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400, fax [phone], or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, 201-748-6011, fax [phone], or online at http://www.wiley.com/go/permission.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at 877-762-2974, outside the United States at 317-572-3993 or fax [phone].

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data:

Thyagarajan, K. S.Still image and video compression with MATLAB / K.S. Thyagarajan.p.cm.ISBN 978-0-470-48416-6 (hardback)1. Image compression. 2. Video compression. 3. MATLAB. I. Title.TA1638.T48 2010006.6′96–dc22

2010013922

oBook ISBN: 978-0-470-88692-2ePDF ISBN: 978-0-470-88691-5

To my wife Vasu,who is the inspiration behind this book

PREFACE

The term “video compression” is now a common household name. The field of still image and video compression has matured to the point that it is possible to watch movies on a laptop computer. Such is the rapidity at which technology in various fields has advanced and is advancing. However, this creates a need for some to obtain at least a simple understanding behind all this. This book attempts to do just that, to explain the theory behind still image and video compression methods in an easily understandable manner. The readers are expected to have an introductory knowledge in college-level mathematics and systems theory.

The properties of a still image are similar to those of a video, yet different. A still image is a spatial distribution of light intensity, while a video consists of a sequence of such still images. Thus, a video has an additional dimension—the temporal dimension. These properties are exploited in several different ways to achieve data compression. A particular image compression method depends on how the image properties are manipulated.

Due to the availability of efficient, high-speed central processing units (CPUs), many Internet-based applications offer software solutions to displaying video in real time. However, decompressing and displaying high-resolution video in real time, such as high-definition television (HDTV), requires special hardware processors. Several such real-time video processors are currently available off the shelf. One can appreciate the availability of a variety of platforms that can decompress and display video in real time from the data received from a single source. This is possible because of the existence of video compression standards such as Moving Picture Experts Group (MPEG).

This book first describes the methodologies behind still image and video compression in a manner that is easy to comprehend and then describes the most popular standards such as Joint Photographic Experts Group (JPEG), MPEG, and advanced video coding. In explaining the basics of image compression, care has been taken to keep the mathematical derivations to a minimum so that students as well as practicing professionals can follow the theme easily. It is very important to use simpler mathematical notations so that the reader would not be lost in a maze. Therefore, a sincere attempt has been made to enable the reader to easily follow the steps in the book without losing sight of the goal. At the end of each chapter, problems are offered so that the readers can extend their knowledge further by solving them.

Whether one is a student, a professional, or an academician, it is not enough to just follow the mathematical derivations. For longer retention of the concepts learnt, one must have hands-on experience. The second goal of this book, therefore, is to work out real-world examples through computer software. Although many computer programming languages such as C, C++, Java, and so on are available, I chose MATLAB as the tool to develop the codes in this book. Using MATLAB to develop source code in order to solve a compression problem is very simple, yet it covers all grounds. Readers do not have to be experts in writing cleaver codes. MATLAB has many built-in functions especially for image and video processing that one can employ wherever needed. Another advantage of MATLAB is that it is similar to the C language. Furthermore, MATLAB SIMULINK is a very useful tool for actual simulation of different video compression algorithms, including JPEG and MPEG.

The organization of the book is as follows.

Chapter 1 makes an argument in favor of compression and goes on to introduce the terminologies of still image and video compression.

However, one cannot process an image or a video before acquiring data that is dealt within Chapter 2. Chapter 2 explains the image sampling theory, which relates pixel density to the power to resolve the smallest detail in an image. It further elucidates the design of uniform and nonuniform quantizers used in image acquisition devices. The topic of sampling using nonrectangular grids—such as hexagonal sampling grids—is not found in most textbooks on image or video compression. The hexagonal sampling grids are used in machine vision and biomedicine. Chapter 2 also briefly demonstrates such a sampling technique with an example using MATLAB code to convert an image from rectangular to a hexagonal grid and vice versa.

Image transforms such as the discrete cosine transform and wavelet transform are the compression vehicles used in JPEG and MPEG standards. Therefore, unitary image transforms are introduced in Chapter 3. Chapter 3 illustrates the useful properties of such unitary transforms and explains their compression potential using several examples. Many of the examples presented also include analytical solutions.

The theory of wavelet transform has matured to such an extent that it is deemed necessary to devote a complete chapter to it. Thus, Chapter 4 describes the essentials of discrete wavelet transform (DWT), its type such as orthogonal and biorthogonal transforms, efficient implementation of the DWT via subband coding, and so on. The idea of decomposing an image into a multilevel DWT using octave-band splitting is developed in Chapter 4 along with examples using real images.

After introducing the useful compression vehicles, the method of achieving mathematically lossless image compression is discussed in Chapter 5. Actually the chapter starts with an introduction to information theory, which is essential to gauge the performance of the various lossless (and lossy) compression techniques. Both Huffman coding and arithmetic coding techniques are described with examples illustrating the methods of generating such codes. The reason for introducing lossless compression early on is that all lossy compression schemes employ lossless coding as a means to convert symbols into codes for transmission or storage as well as to gain additional compression.

The first type of lossy compression, namely, the predictive coding, is introduced in Chapter 6. It explains both one-dimensional and two-dimensional predictive coding methods followed by the calculation of the predictor performance gain with examples. This chapter also deals with the design of both nonadaptive and adaptive differential pulse code modulations, again with several examples.

Transform coding technique and its performance are next discussed in Chapter 7. It also explains the compression part of the JPEG standard with an example using MATLAB.

The wavelet domain image compression topic is treated extensively in Chapter 8. Examples are provided to show the effectiveness of both orthogonal and biorthogonal DWTs in compressing an image. The chapter also discusses the JPEG2000 standard, which is based on wavelet transform.

Moving on to video, Chapter 9 introduces the philosophy behind compressing video sequences. The idea of motion estimation and compensation is explained along with subpixel accurate motion estimation and compensation. Efficient techniques such as hierarchical and pyramidal search procedures to estimate block motion are introduced along with MATLAB codes to implement them. Chapter 9 also introduces stereo image compression. The two images of a stereo image pair are similar yet different. By using motion compensated prediction, the correlation between the stereo image pair is reduced and hence compression achieved. A MATLAB-based example illustrates this idea clearly.

The concluding chapter, Chapter 10, describes the video compression part of the MPEG-1, -2, -4, and H.264 standards. It also includes examples using MATLAB codes to illustrate the standards′ compression mechanism. Each chapter includes problems of increasing difficulty to help the students grasp the ideas discussed.

I thank Dr. Kesh Bakhru and Mr. Steve Morley for reviewing the initial book proposal and giving me continued support. My special thanks to Dr. Vinay Sathe of Multirate Systems for reviewing the manuscript and providing me with valuable comments and suggestions. I also thank Mr. Arjun Jain of Micro USA, Inc., for providing me encouragement in writing this book. I wish to acknowledge the generous and continued support provided by The MathWorks in the form of MATLAB software. This book would not have materialized but for my wife Vasu, who is solely responsible for motivating me and persuading me to write this book. My heart-felt gratitude goes to her for patiently being there during the whole period of writing this book. She is truly the inspiration behind this work.

K. S. THYAGARAJAN