A novel self-organizing clustering network and its application in data fusion for target classification

A novel selforganizing clustering artificial neural network (DIGNET) and its unsupervisedlearning algorithm are introduced. In view of the characteristics of data fusion and target recognition,a DIGNET based data fusion approach at decision level is proposed. The clustering performances of DIGNET and selforganizing feature mapping network (SOFM) and proposed data fusion architecture are studied using simulated data. The experimental results show that the correct classification rate of DIGNET and its tolerance to noise interference are high, compared with SOFM; the fusion recognition can be effectively carried out with the DIGNET based data fusion system. The fusion architecture is used in a target tracing system of FLIR and TV camera and the results indicate that it is viable for practical purpose.