The Use of High Performance Computing in JPEG Image Compression

The problem inherent to any digital image (or digital video) system is the large amount of bandwidth required for transmission or storage. This has driven the research area of image compression to develop more complex algorithms that compress images to lower data rates with better fidelity. One approach that can be used to increase the execution speed of these complex algorithms is to implement the algorithms on a parallel supercomputer. In this paper we address the parallel implementation of the JPEG still image compression standard on the MasPar MP-1, a massively parallel SIMD computer. We demonstrate that the greatest difficulty lies not with the compression algorithm per se, but with the speed bottleneck that arises in the output of the compressed image. We develop a parallel output algorithm which addresses this problem and present results which show real-time performance on 1024× 1024 images.