A MICROCOMPUTER-BASED IMAGE PROCESSOR FOR AUTOMATIC INDEXING OF PRINTED CIRCUIT BOARDS

Abstract

Some steps in the assembly of Printed Circuit Boards (PCBs) have not been successfully automated. For example, design modifications in the form of jumper wires are often manually applied to the boards as a final step in the production. To automate this process, the locations of the points of interest on the boards must be known precisely. The traditional approach to this problem involves precise mechanical alignment of the boards through the use of indexing pins or clamps. The disadvantage of this method lies in the expense of a mechanical fixturing device with the required tolerances. As an alternative method, this thesis describes the development and test results of a microcomputer-based system designed to locate features on the surface of a printed circuit board through techniques of image processing, in particular, image registration. The system is a combination of hardware and software that has the ability to acquire, process and store images. Control and monitoring of results are achieved through human interaction with the system. Tests on the system reveal that rapid registration of printed circuit board images by microcomputer-based hardware is practical. For strictly translational offsets, registration accuracy to one pixel was achieved in times on the order of 1.8 seconds. Estimates of the rotational offsets can also be derived.