Parallel neural network recognition - a multi-agent system approach

Abstract

This paper presents and discusses the design and the development of a pattern recognition agent based on neural networks. This agent is part of an intelligent navigation system, providing it with the necessary vision abilities so that it can be placed on a strange environment in order to explore and recognise its structures and specificities. Although similar, the properties of the recognised objects change through time and according to each specific environment. The flexibility required by such recognition process was implemented by several pattern recognition agents. Each agent is based on a neural network and can be trained on-line by a parallel training algorithm to allow an effective real time utilisation.