SYSTEM AND METHOD FOR VARIABLE ENCODING BASED ON IMAGE CONTENT

A system and method for efficient image encoding begins by loading a series of image blocks. A current block is compared with an immediately preceding image block for coincidence, and coded as matching or nonmatching based on the comparison results. The system further compares the nonmatching image blocks with a stack containing a list of the most recently encountered image blocks. Image blocks matching an entry in the list are coded by reference to the appropriate entry in the list. In the event a match is still unavailing, the system determines if the image blocks may be categorized as a bilevel text block, a bilevel image block, a block containing only one gray pixel, or a block containing multiple gray pixels. Each nonmatching block is classified according this classification criteria.