Optimizing fusion architectures for limited training data sets
Brian A. Baertlein, Ajith H. Gunatilaka · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2000
A method is described to improve the performance of sensor fusion algorithms. Data sets available for training fusion algorithms are often smaller than described, since the sensor suite used for data acquisition is always limited by the slowest, least reliable sensor. In addition, the fusion process expands the dimension of the data, which increases the requirement for training data. By using structural risk minimization, a technique of statistical learning theory, a classifier of optimal complexity can be obtained, leading to improved performance. A technique for jointly optimizing the local decision thresholds is also described for hard- decision fusion. The procedure is demonstrated for EMI, GPR and MWIR data acquired at the US Army mine lanes at Fort AP Hill, VA, Site 71A. It is shown that fusion of features, soft decisions, and hard decisions each yield improved performance with respect to the individual sensors. Fusion decreases the overall error rate from roughly 20 percent for the best single sensor to roughly 10 percent for the best fused result.