A Novel Fuzzy Pattern Recognition Data Association Method for Biased Sensor Data

Data association is one of the most important problems in multiple-sensor multiple-target tracking systems. A novel approach for biased-data association based on patterns extracted from topologies of measurements is presented in this paper. The introduction of patterns reveals a new paradigm for exploiting redundant spatial information in association. Fuzzy pattern recognition method has been applied to analyze the similarity of target patterns, and the criterion of association can be set up accordingly. By virtue of the inherent character of patterns, the proposed pattern-based association algorithm is robust to large registration errors between systems. Furthermore, the association results can provide feedback information for sensor registration. Simulation results show that the proposed algorithm is effective and feasible in solving association problems under the conditions of sensor bias