Abstract: An input frame of video sequence is partitioned into smaller blocks of pixel data where each block is subjected to an edge detection which examines the activities in the block itself and the surrounding blocks to determine whether the block can be classified as an edge block. If the block is classified as a non-edge block, the conventional DCT coding method (e.g. SM3) with or without motion compensation is applied to the block. On the other hand, if the block is classified as an edge block, the block is subjected to a pixel value prediction process with or without reference to the previously coded frames, followed by a quantization process, and a run-length encoding process with variable length codes, i.e., a DPCM process.