Abstract
We analyze statistical encoding algorithms as one of the type of general methods of image compression. An approach we proposed allows us to improve effectiveness of a lossy image compression. We develop a statistical encoding algorithm that remains effective when compressing images with raw errors. It can be used as a part of any methods of image compression performing encoding of decorrelated signals with nonuniform distribution of probabilities. Coding specific data of a hierarchical compression method, we experimentally compared our algorithm with ZIP and ARJ archivers.
