Introduction to Data Compression (Morgan Kaufmann Series in Multimedia Information and Systems) - Hardcover

Sayood, Khalid

Hardcover

ISBN 10: 1558603468 ISBN 13: 9781558603462

Publisher: Morgan Kaufmann Pub, 1996

Synopsis

This important new multimedia book concentrates on the significance of data compression for storing and transmitting large masses of information for all media types. Algorithms, examples, and discussions of official data compression standards for each medium allow practicing engineers and computer scientists to efficiently manipulate these growing masses of information.

About this title

Review

Khalid Sayood's textbook-style Introduction to Data Compression is the definitive guide to all kinds of compression schemes. Early chapters establish the mathematics involved in basic compression techniques, including lossless and lossy compression as well as the fundamentals of information theory that lay the groundwork for common forms of compression. (The book contains all the relevant formulas, although those who don't need such mathematical detail will still be able to understand the book.)

A good portion of the book examines various compression schemes, their strengths and weaknesses, and what content they work best for. Introduction to Data Compression begins with lossless compression schemes, which lose no information during the compression/decompression process. Huffman Coding, a well-established compression scheme, and arithmetic and dictionary coding also receive excellent treatment. In addition, the author takes on lossless compression for images.

For lossy compression, Sayood discusses schemes that use quantization, where a range of values is compressed in some way. He also describes scalar, vector, and differential encoding and fractal compression. A final chapter looks at video encryption (which often combines techniques from earlier chapters). Many of the compression schemes include examples from image and sound files, but the book considers a wide variety of video schemes too. This rich and confidently written text collates a lot of research and can serve as both textbook and source for designers who need a readable and mathematically solid introduction to data compression.

Review

The need to efficiently store, manipulate, and transmit large masses of information is growing more rapidly than the capacity of systems to handle it. Engineers and computer scientists need a solid understanding of compression in order to work with the burgeoning variety of data types and increasingly data-intensive applications. This uniquely comprehensive book explains the fundamental theories and techniques of data compression, with the most complete coverage available of both lossy and lossless methods. Rather than simply describing current approaches, Sayood explains the theoretical underpinnings of the algorithms so that readers learn how to model structures in data and design compression packages of their own. Practitioners, researchers, and students will benefit from the balanced presentation of theoretical material and implementations.

Features:
* Covers both lossy and lossless compression techniques with applications to image, speech, text, audio, and video compression.
* Official compression standards for video, audio, text, and facsimile are discussed in order to illustrate the techniques: includes JPEG, MPEG, G.728, H.261, and Group 3 and 4 fax standards.
* Detailed examples follow each new concept or algorithm.
* Software implementations and sample data sets are available, allowing readers to work through the examples in the book and to experiment with various compression techniques on their own.
* Optional starred sections provide enhanced technical or theoretical discussions.
* Appendices on probability theory, random processes, and matrix concepts are included for reference. -- Book Description

Edition details

• Publisher: Morgan Kaufmann Pub
• Publication date: 1996
• ISBN 10: 1558603468
• ISBN 13: 9781558603462
• Binding: Hardcover
• Language: English
• Edition number: 2
• Number of pages: 475