Architectural approaches for video compression

Abstract

An overview on architectures for implementations of current video compression schemes is given. Dedicated as well as programmable approaches are discussed. Examples for dedicated function-specific implementations include architectures for DCT and block matching. For programmable video signal processors, a number of architectural measures to increase video compression performance are reviewed. Actual implementations of video compression schemes typically employ a variety of different architectural approaches. The detailed mix of approaches depends on the targeted application spectrum.