In view of the features of multi-sensor data fusion and target recognition, a modified counter-propagation neural network (MCPN) is proposed. The competitive layer of the MCPN is based on Dignet algorithm but not Kohonen clustering network (KCN). And the MCPN based fusion architecture at decision level is presented and applied to multi-sensor data fusion for target classification. The proposed MCPN and the fusion architecture are studied using simulated data. The experimental results show that the fusion classification can be effectively realized with the MCPN compared with standard counter-propagation neural network (CPN) and back-propagation network (BPN) of the same size. Finally, to further illustrate its effectiveness for practical uses, the experiments are conducted using real-world data acquired with a target tracing system of FUR and TV camera. The results indicate that the MCPN and its fusion architecture are workable

Date Published: 16 September 2002
PDF: 6 pages
Proc. SPIE 4929, Optical Information Processing Technology, (16 September 2002); doi: 10.1117/12.483204

Published in SPIE Proceedings Vol. 4929:
Optical Information Processing Technology
Guoguang Mu; Francis T. S. Yu; Suganda Jutamulia, Editor(s)