Distributed Image Edge Detection Methods and Their

Communication Pattern Implications ?

Xiaodong Zhang Sandra G. Dykes
High-Performance Computing and Software Laboratory
The University of Texas at San Antonio
San Antonio, Texas 78249

Hong Deng
Diversified Technology Inc.
San Antonio, Texas 78247

Abstract

An edge detection process in computer vision and image processing detects any types of significant features appearing as discontinuities in intensities. This paper presents our experience with parallelizing an edge detection application algorithm that reduces noise and unnecessary detail in a gray-scale image from a coarse level to a fine level of resolution by using an edge focusing technique. Numerical methods and parallel implementations of edge focusing are presented. The image detection algorithms are implemented on three representative massage-passing architectures: a low-cost heterogeneous PVM network, an Intel iPSC/860 hypercube, and a CM-5 massively parallel multicomputer. The CM-5 studies include both message-passing and dataparallel versions. Our objectives are to provide insight into implementation and performance issues for image processing applications on general-purpose message-passing architectures, to investigate implications on network variations, and to evaluate the computing scalabilities on the three network systems by examining execution and communication patterns of the image edge detection application.
?This work has been supported in part by the National Science Foundation under grants CCR-9102854 and CCR-9400719, by the U.S. Air Force under Agreement FD-204092-64157, by Air Force Office of Scientific Research under grant AFOSR-95-1-0215, and by a grant of Cray Research. The parallel experiments were performed on the Intel iPSC/860 machines in the Center for Research on Parallel Computation at Rice University, and in the Intel Supercomputer System Division at Portland, Oregon; and on the CM-5 machines in Los Alamos National Laboratory, and in the National Center for Supercomputing Applications at the University of Illinois.