An efficient architecture for the in place fast cosine transform

The cosine transform (DCT) is in the core of image encoding and compression applications. We present a new architecture to efficiently compute the fast direct and inverse cosine transform which is based on reordering the butterflies after their computation. The designed architecture exploits locality, allowing pipelining between stages and saving memory (in place). The result is an efficient architecture for high speed computation of the DCT that reduces significantly the area required to VLSI implementation.