A novel algorithm for reducing computational complexity of MC-DCT in frequency-domain video transcoders

摘要

In order to transmit pre-encoded digital video over heterogeneous networks, it becomes necessary to employ transcoding techniques that convert pre-encoded video streams into streams having different bit rates and quality. This process is referred to as rate shaping or rate adaptation. Such video transcoders save computational overheads incurred in the IDCT (stream decode) and DCT (stream encode) operations, by performing motion compensation (MC) in the frequency (block-DCT) domain (MC-DCT). MC-DCT involves pre- and post-multiplication of 8×8 DCT blocks with 8×8 matrices derived from motion vectors. The specific contribution of this work is towards developing a novel method for implementing the pre- and post-multiplication with minimum computational overhead. The proposed method gives time savings of up to 50% over normal block multiplication.