A NEURAL NETWORK RECOGNITION SCHEME FOR CLASSIFYING TWO AND THREE-DIMENSIONAL INDUSTRIAL WORKPIECES BASED ON THE EDGE DETECTION CONTOUR SEQUENCE

In automated visual industrial inspection, computer vision systems have been widely used. Such systems are often application specific, and therefore require domain knowledge in order to have successful implementation. Since visual inspection can be viewed as a decision making process, it is argued that the integration of artificial neural networks and computer vision systems provides a practical approach to general purpose visual inspection applications. As a consequence, the research work presented in this paper investigates several novel uses of machine vision and artificial neural networks in the processing of single camera multi-positional images for two and three-dimensional object recognition and classification. Many industrial applications of machine vision allow objects to be identified and classified by their boundary contour or silhouette. Boundary contour information was chosen as an effective method of representing the industrial component, a composite signature being generated using Euclidean arrays obtained from the generation of multi-centroidal positions and the boundary pixels.