Video based Object Recognition for Conveyor Line Automation

The aim of this book is to introduce a vision based object recognition system for automation of assembly line. The vision system is designed to recognize the objects, which are placed on the assembly line and also to identify the exact position for tracking each particular object, in order to lead the robot manipulator system for further processing. The method consists of an object shape discrimination analysis methodology using image processing techniques for the video frames acquired through the CCD camera over the conveyor line. The recognition of objects have been tested and evaluated using three different feature extraction methods on twenty objects for model conveyer line. These methods utilize geometric features, two dimensional moments and one dimensional scaled normalizing moments. The one dimensional scaled normalized moment method has been reported to be the best one for the object recognition. The neural network classifier is constructed for the features using the theoretical principles of artificial neural networks (ANN). Furthermore a step-by-step methodology for designing a neural network structure for the proposed system has been introduced.