Method and associated apparatus for encoding bit planes.

摘要

An encoding method is described for image bit planes whereby an arithmetic coder using conditioning contexts based on pixels from both the current bit plane and previously decoded bit planes is employed. For the same number of contexts (which is important for minimizing memory usage), the method provides for higher compression than those methods based on a single bit plane context. An additional improvement over the prior art is that the method eliminates the need to form a Gray code bit plane representation, as is commonly done to reduce the correlation among bit planes. However, if used on conjunction with the gray code, a small improvement in compression performance will still be realized. <IMAGE>