PARALLEL IMAGE MATCHING ON A DISTRIBUTED SYSTEM

Image matching based on image feature pixels involves heavily iterated computation and frequent memory access. The key to increase the speed is to employ parallelism on either parallel machines or workstation c lusters. This paper presents the development of a parallel image match ing system which uses a divide-and-conquer method to implement the pro posed hierarchical matching scheme on a networked workstation cluster. Our investigation shows that a distributed workstation cluster can be st meet the demand of high computation and memory access in image proc essing. The performance or our proposed matching scheme is evaluated i n terms of execution time.