Chapter 5 - Arithmetic Coding

Abstract

In this chapter, arithmetic coding is divided into two parts, the first part explains how and why arithmetic coding works, and the second part deals with some of the common properties that are later used for computational techniques required for a practical implementation. Compression applications use a wide variety of techniques and have quite different degrees of complexity, but also share some common processes. Numerical processing, like predictive coding and linear transforms, is used normally for waveform signals such as images and audio. The next stage, source modeling, is used to account for variations in the statistical properties of data. It is responsible for gathering statistics and identifying data contexts that make the source models accurate and reliable. Arithmetic coding stands out as one method that is able to work most efficiently in the largest number of circumstances and purposes, and it also stands out in terms of elegance, effectiveness, and versatility. However, with all these advantages, arithmetic coding is not as well understood as the other methods. Certain practical problems hold back its adoption. However, most of these issues have been tackled, and the relative efficiency of computer arithmetic has improved and new techniques have made it cost-effective to operate. As most of the patents have expired or become obsolete, current computational resources make it possible to implement simple, efficient, and royalty-free arithmetic coding.