A DCT-Domain Video Transcoder for Spatial Resolution Downconversion

Abstract

Video transcoding is an efficient way for rate adaptation and format conversion in various networked video applications. Several transcoder archi-tectures have been proposed to achieve fast processing. Recently, thanks to its relatively low complexity, the DCT-domain transcoding schemes have become very attractive. In this paper, we investigate efficient architectures for video downscaling in the DCT domain. We propose an efficient method for composing downscaled motion vectors and determining coding modes. We also present a fast algorithm to extract partial DCT coefficients in the DCT-MC operation and a simplified cascaded DCT-domain video transcoder architecture.