Tactile Sensor Array Signal and Data Processing Based on Information Fusion

Abstract

It developed a sort of tactile sensor array system that built in various shapes and provided a wide range of spatial resolution. The sensor system built a demonstration system in lab and developed algorithms to process the large amount of tactile data. The efficient fusion of data from different sensors would enable the robot to respond promptly in dealing with the "real world". The confidence distance measure was used as fusion measure of information fusion. The useful fused data found by confidence distance matrix and relation matrix. Finally, the optimal fused data was given by maximum principle and maximum likelihood method. Using algorithm based information fusion, the influence of single sensor cell uncertainty errors could be avoid, surface depth data measure and image reconstruction precision of object profile could be improved.