Combining the Opinions of Several Early Vision Modules using a Multi-Layer Perceptron

This paper deals with the solution of a binary classification problem by acting on the combined evidence of several early vision modules. Each module provides an opinion on the identity of an individual image element based on a specific area of expertise, such as texture, motion, depth etc. The problems involved in reaching a consensus of opinion are discussed, and the effectiveness of using a trained Multi-Layer Perceptron as a tool for data fusion is examined. Some preliminary results are reported. The original version of the document has been extended with new Appendices to cover recent developments in the field of pattern recognition and machine learning. We hope to show with this additional material that this approach to solving the data fusion problem is just as valid now as when we first suggested it, and does not have the failings of more recent and quite widespread approaches. Research Initiative in Pattern Recognition, DRA Electronics Division, St. Andrews Road, Malvern, Worcs. AI Vision Research Unit, University of Sheffield, Sheffield, Yorks. Computing and Maths Dept., Oxford Polytechnic, Oxford.