Intelligent Robotic Vision Systems

Abstract

This chapter covers the current weaknesses and the directions of current research. It focuses on all relevant aspects, which means that vision computers are treated as part of an environment that includes feeding systems, robots, light sources, and sensors. The chapter describes both software and hardware for image processing. It deepens the interactions of the vision system with the robot and the part transport systems. The chapter provides an overview of the different aspects of a vision system. It also describes the different steps of choosing a good illumination technique, selecting the appropriate type of sensor, and then defining the spatial sampling and gray-level quantization characteristics and selecting the vision computer. The chapter concentrates on algorithms for the three major tasks of robot vision: object recognition and localization, visual servoing, and visual inspection. The reader should be familiar with basic image-processing techniques, such as convolution, Fourier transforms, thresholding, and run length encoding.