| Compression schemes can be divided into two major classes: lossless compression schemes and lossy compression schemes. Data compressed using lossless compression schemes can be recovered exactly, while lossy compression introduces some loss of information in the reconstruction.
While the first modern compression scheme, Huffman coding, was a lossless compression scheme, most of the initial activity in the compression area focused on lossy compression. One reason was that the "data" being considered for compression was analognprimarily samples of speech waveforms. Another reason was that the device which would be used to provide the reconstructed speech to the user, namely, the telephone, introduced so much distortion of its own that it was relatively easy to design compression schemes which introduced less distortion than that to which the user was already being subjected.
In the past two decades the situation has changed dramatically. There is a significant amount of discrete data in the form of text, graphics, images, video, and audio that needs to be stored or transmitted, and display devices are of such quality that very little distortion can be tolerated. This has resulted in a resurgence of interest in lossless compressionnhence this handbook. We have tried to put together, in an accessible form, some of the most important aspects of lossless compression. The idea was to have a volume which would allow the reader to get an idea of both the depth and the breadth of the field while at the same time fulfilling the usual function of a handbookmthat of being a convenient repository of information about a topic. This combination will, we hope, be useful to both the novice reader who wishes to learn more about the subject, as well as the practitioner who needs an accessible reference book. |
