ABSTRACT
Of crucial importance to the success of image compression is the concept of redundancy, which denotes the predictability of spatial, statistical, or motion parameters derived from images or sequences of images. This entry begins with a definition of natural imagery and remarks concerning the occurrence of common redundant patterns in nature. Redundancy at the level of pixel values is discussed, and the types and examples of spatial patterns in natural and man-made objects that facilitate compression are considered. Also, an overview of redundancy in transformed pixel values is provided, and analytical and empirical measures of redundancy and their computation are discussed.
