A low-complexity frame memory compression algorithm and its implementation for MPEG-2 video decoder

An algorithm and its implementation to compress video frame data to be stored in external memory are proposed, thereby enabling a reduction of memory requirements and bandwidth. To support random access capability, frames are organized as 1/spl times/8 fixed-length segments, which are used as compression units. For both compression efficiency and low complexity, modified Hadamard transform (MHT) and Golomb-Rice (GR) coding are utilized. The compression ratio is fixed at 50% and objective and subjective video quality assessments are performed. Efficient ASIC implementation including Golomb-Rice encoding/decoding is also proposed to reduce the total compression/decompression latency time.