Robot Vision

From the Publisher: This book presents a coherent approach to the fast-moving field of computer vision, using a consistent notation based on a detailed understanding of the image formation process. It covers even the most recent research and will provide a useful and current reference for professionals working in the fields of machine vision, image processing, and pattern recognition. An outgrowth of the author's course at MIT, Robot Vision presents a solid framework for understanding existing work and planning future research. Its coverage includes a great deal of material that is important to engineers applying machine vision methods in the real world. The chapters on binary image processing, for example, help explain and suggest how to improve the many commercial devices now available. And the material on photometric stereo and the extended Gaussian image points the way to what may be the next thrust in commercialization of the results in this area. Chapters in the first part of the book emphasize the development of simple symbolic descriptions from images, while the remaining chapters deal with methods that exploit these descriptions. The final chapter offers a detailed description of how to integrate a vision system into an overall robotics system, in this case one designed to pick parts out of a bin. The many exercises complement and extend the material in the text, and an extensive bibliography will serve as a useful guide to current research. Errata (164k PDF)