作者: Angshuman Saha , Satish Soman
DOI:
关键词:
摘要: A video decoder is provided with robust error handling and concealment. In one embodiment, the detects syntactic, semantic, coding errors in encoded slices of macroblocks. An handler determines number remaining un-decoded macroblocks corrupted slice replaces these using substitute DCT coefficient matrices motion vectors. The zero-frequency each matrix set equal to last uncorrupted macroblock, while higher frequency coefficients are zero. vectors from a concealment vector memory which buffers previous macroblock row. this way, intelligent approximations made for missing macroblocks, effectively masking bitstream error.