Reducing Memory In MPEG2-video-decoder-architecture

Abstract

Different possibilities of reducing the memory needed in MPEG2-video-decoder-architectzrres are shown and discussed. The memory of an MPEG-decoder consists of the input buffer, the reference picture memories and apicture output buffer used for progressive to interlaced conversion. The reference picture memories and the picture output buffer are considered for reduction. The most promising approach of reducing the reference picture memories requires twice the calculation power of the usual approach and will show signifxant memory savings for the higher levels of MPEGZ.