DCT-based video frame-skipping transcoder

摘要: Due to high computational complexity and quality degradation introduced by conventional frame-skipping transcoders, a DCT-based video transcoder was proposed. However, the transcoding process of motion compensated macroblocks in DCT domain becomes bottleneck since IDCT processes are required. In this paper, we propose new architecture reduce process. The transcodes dominant region an MC macroblock making use coefficients from incoming bistream some pre-computed shifting operators. By using our proposed shifted version vector, re-encoding errors can be avoided. On other hand, adaptive transcode boundary marcoblocks perform error compensation is also Experimental results show that, as compared transcoder, for more robust noise, gives rise fewer requantization errors, requires simple complexity.