Bayesian classification for inspection of industrial products

Abstract

In this paper, a real time application for visual inspection and classification of cork stoppers is presented. Each cork stopper is represented by a high dimensional set of characteristics corresponding to relevant visual features. We have applied a set of non-parametric and parametric methods in order to compare and evaluate their performance for this real problem. The best results have been achieved using Bayesian classification through probabilistic modeling in a high dimensional space. In this context, it is well known that high dimensionality does not allow precision in the density estimation. We propose a Class-Conditional Independent Component Analysis (CC-ICA) representation of the data that even in low dimensions, performs comparably to standard classification techniques. The method has achieved a success of 98% of correct classification. Our prototype is able to inspect the cork stoppers and classify in 5 quality groups with a speed of 3 objects per second. © Springer-Verlag Berlin Heidelberg 2002.