Macropipelined Multicomputer Architecture for Image Analysis.

摘要

A scheme is presented for macropipelining an L-stage image analysis algorithm into a message-passing N, processor multicomputer system in which N, > L. The resulting architectures achieve high speeds in processing multiple images. The macropipelining scheme is tested on a target recognition algorithm in a simulated hypercube computer system. The results support the belief that this scheme yields effective architectures for high-speed processing of long sequences of images. Keywords: Multicomputers; Computer architecture; Reprints. (jhd)