A novel VLSI architecture for image compression

Abstract

In Analog signal processing of multimedia applications, audio need not be compressed but video must be compressed in order to conserve the bandwidth. For this purpose Fast Cosine Transform (FCT) is preferred, which is developed using Fast Fourier Transform (FFT). But it has several limitations of Delays, Area and Power. In this paper the similarities among Fast Fourier Transform (FFT), Fast Cosine Transform (FCT) and Fast Hartley Transform (FHT) have been studied. The Converged Very Large Scale Integration (VLSI) architecture for these Hybrid Transforms for video compression are designed, simulated and synthesized. Implementing FCT, FFT, and FST from Fast Hartley Transform is developed and Hybrid Transforms architecture is proposed. Further more, delays and area overheads have been calculated. For all these transforms the layouts have been drawn using Magma Tools with 0.13(im technology. Compression ratio for image with each transform and Hybrid transform is also implemented © 2006 IEEE.