The GMCPHD Filter for Irregular Group Target Spawning Based on Star-Convex RHMs

Target spawning and extended shape estimation are important problems in group target tracking. In this paper, we propose a Gaussian mixture cardinalized probability hypothesis density (GMCPHD) filter for group targets with spawning and irregular shape based on star-convex Random Hypersurface Model (RHM). In order to solve the problem of irregular group shape, we use star-convex RHM to describe the distribution of measurement sources. Besides, we use the distance division method to realize the division of measurement sets and the judgment of group splitting. On this basis, the real-time tracking of the motion state and extended shape is realized under the framework of GMCPHD. The performance of this algorithm is showcased by comparison with the elliptical RHM-based GMCPHD filter, and the results show that the proposed algorithm can improve the estimation accuracy of group shape and motion state effectively.