AN ON-LINE IMAGE PRE-PROCESSING CAMERA SYSTEM FOR PATTERN RECOGNITION.

摘要

A pattern recognition system has been developed which is capable of recognizing high contrast two-dimensional visual patterns and which is insensitive to the translation, rotation, and size of a pattern. The unique characteristics of the system are its speed, its low cost and the fact that it is completely self-contained. The system is intended primarily for industrial robotics applications.;The camera system is used to input an image, to center it, to normalize it with respect to size, and to convert it into polar coordinate form. The input device is a standard closed-circuit television camera. Rough centering is performed electro-mechanically and fine centering is done electronically. Size normalization is accomplished by utilizing a motorized zoom lens.;Four experiments were performed to test the camera system. The results of these experiments are discussed and possible methods of increasing the speed of the system still further are suggested.;The basic idea of this pattern recognition system is to do as much processing as possible at the front end of the system, thereby presenting the pattern recognition unit with pre-processed data which can be analyzed quickly. The project was thus broken down into an image pre-processing camera system and a pattern recognition unit. This thesis describes the camera system.