SHARP - A SHAPE-RECOGNITION SYSTEM AND ITS PARALLEL IMPLEMENTATION

A parallel algorithm for shape recognition is presented along with its implementation on a distributed memory multiprocessor. Shape recognit ion is one of the fundamental problems of computer vision. We consider a shape to be composed of a set of small straight line segments tange ntial to the object. The recognition problem is to determine whether t he test image contains a specified reference shape or not. The straigh t line Hough transform [SLHT] has been used to detect reference shapes . A signature-based parallel algorithm called SHARP is developed for s hape recognition using SLHT on a distributed memory multiprocessor sys tem. in the SHARP algorithm, the (theta, r) space is divided among pro cessors. The SHARP algorithm has been implemented on a Meiko transpute r with 32 nodes. We analyse the performance of the parallel algorithm using both theoretical and experimental techniques.