Vision and Information Processing for Automation

Developments in electronic hardware, particularly microprocessors and solid-state cameras, have resulted in a vast explosion in the range and variety of applications to which intelligent processing may be applied to yield cost-effective automation. Typical examples include automated visual inspection and repetitive assembly. The technology required is recent and specialized, and is thus not widely known. VISION AND INFORMATION PROCESSING FOR AUTOMATION has arisen from a short course given by the authors to introduce potential users to the technology. Its content is a development and extension of material presented in the course. The objective of the book is to introduce readers to modern concepts and techniques basic to intelligent automation, and explain how these are applied to prac tical problems. Its emphasis is on machine vision. Intelligent instrumentation is concerned with processing infor mation, and an appreciation of the nature of information is essential in configuring instrumentation to handle it effiCiently. An understand ing of the fundamental principles of efficient computation and of the way in which machines make decisions is vital for the same reasons. Selection of appropriate sensing (e.g., camera type and configuration), of illumination, of hardware for processing (microchip or parallel processor?) to give most effective information flow, and of the most appropriate processing algorithms is critical in obtaining an optimal solution. Analysis of performance, to demonstrate that requirements have been met, and to identify the causes if they have not, is also important. All of these topics are covered in this volume.