Pattern recognition, image processing and computer vision in fifth generation computer systems

Abstract

It is well-known that one of the goals of research for the last two decades or so in pattern recognition and its sub-areas, such as processing, analysis and understanding of image, speech and natural language, and computer vision techniques etc., has always been to develop fundamental techniques for flexible interactive intelligent man-machine interfaces for computers. In this paper, the author tries to argue that for the evolution of fifth generation computer systems (FGCS) as defined by Japanese scientists, some of the things required are realisation and implementation of the advances in pattern recognition and its sub-areas, not only to achieve the man-machine interface with a natural mode of communication, but also for the realisation of the basic mechanisms of inference, association and learning, which are inherent both in pattern recognition and in the core functions of FGCS. The next generation computers will be knowledge-based systems, which form a subdomain of artificial intelligence (a1) techniques, and soa1 provides the essential link between pattern recognition domains and different application systems. No attempt is made to discuss other essential conceptual building blocks, such as software engineering, computer architecture and very large scale integration technology unless these become very relevant in the discussions of concerned topics of the paper. A section on limitations of perception, learning and knowledge for computing machines is also included.