Track Correlation Algorithm Based on Multi-dimension Assignment and Gray Theory

Wei Xiong,Jing-wei Zhang,You He
DOI: https://doi.org/10.3724/SP.J.1146.2009.00394
2010-01-01
Abstract:To resolve the problem of track correlation in distributed multi-sensor system,a track correlation algorithm based on multi-dimension assignment and gray theory is presented in this paper. Firstly,gray theory is applied to acquire a global statistical vector in the distributed multi-sensor system. Then,a multi-dimension gray similar degree matrix is build according to the global statistical vector. Based on this matrix,the track correlation results can be got by a multi-dimension assignment method. At last,the algorithm is compared with gray track correlation algorithm. The simulation results show that the performance of the algorithm here is much better than that of the gray track correlation algorithm in dense multi-target environments,more cross,split and maneuvering track situations. In this situations,its correct correlation rate is improved about 8.8 percent over that of the gray track correlation algorithm.
What problem does this paper attempt to address?