TO ACHIEVE A HIGHER LEVEL OF INTELLIGENCE IN MACHINE VISION FOR FACTORY AUTOMATION, SOME FORM OF PARALLEL PROCESSING MUST BE USED. THERE ARE TWO CLASSES OF PARALLELISM WITH WIDESPREAD APPLICATIONS IN INDUSTRY: MASSIVELY PARALLEL PROCESSORS AND THE PIPELINE PROCESSOR. SPEED AND FLEXIBILITY OF THESE TWO TYPES OF ARCHITECTURES ARE COMPARED. THE ONLY MASSIVELY PARALLEL SYSTEM WITH A LOW ENOUGH COST FOR INDUSTRY HAS DATA COMMUNICATION BETWEEN PROCESSORS IN A LINEAR CHAIN. A BRIEF REVIEW OF ALGORITHM TECHNIQUES SHOW HOW IT CAN BE PROGRAMMED FOR A NUMBER OF SOPHISTICATED CONCEPTS, SUCH AS THE 2-D FFT, THE HOUGH TRANSFORM, AND NEURAL NETWORKS.
| Number of Pages : | 12 |
| Published : | 06/01/1990 |
