A genetic-algorithm-based fusion system optimization for 3D image interpretation

Abstract

Information fusion systems are complex systems with many parameters that must be adjusted to obtain interesting results. Generally applied in specialized domains such as military, medical and industrial areas, these systems must work in collaboration with the experts of the domains. As these end-users are not specialists in information fusion, the parameters adjustment becomes a difficult task. In addition, to find a good set of those parameters is a hard and time consuming process as the search space is very large. In order to overcome this issue a genetic algorithm is applied to automatically search the best parameter set. The results show that the proposed approach produces accurate levels of the global performance of the fusion system. © 2010 Springer-Verlag.