HIGH-SPEED IMAGE ROTATION IN EMBEDDED SYSTEMS

The design of an enhanced SIMD processor, named the SIMDDE, is evaluat ed. The architecture has circular data paths, as well as row and colum n paths. The SIMDDE architecture is intended primarily for use in imag e recognition in embedded systems. The designed processor also might b e realized with other architectures provided they meet the spatial con straints. Performance comparisons are made with a von Neumann, an unen hanced SIMD, two pipeline architectures, and a paracomputer. The metho dology employs an image recognition algorithm introduced by Castleman (Digital Image Processing, Prentice-Hall, Englewood Cliffs, NJ, 1979) and an extension of the published analysis of Cantoni, Guerra, and Lev ialdi (Computing Structures for Image Processing (M. Duff, Ed.), pp. 4 3-56, Academic Press, New York, 1983). It is shown that this augmentat ion can increase the performance threefold at only a 6% increase in th e cost as measured by the number of paths required. Comparisons of the design as realized with other architectures indicate when these might be preferred. (C) 1995 Academic Press, Inc.