Improved Cardinalized Probability Hypothesis Density Based on DBSCAN for Multiple Group Targets Tracking

To more accurately track multiple group targets in scenarios involving group conformity and group division, an improved cardinalized probability hypothesis density (ICPHD) based on density-based spatial clustering of applications with noise (ICPHD-DBSCAN) for multiple group targets tracking is proposed. The proposed algorithm addresses the issue of insufficient estimation accuracy due to the underutilization of measurement number information in traditional methods by offering new analytical solution. It considers the impact of measurement number on the group state of targets, effectively enhancing the accuracy and practicality of the estimation. Furthermore, to tackle the exponential partitioning problem that may arise in the GGIW-CPHD filter when dealing with a large number of measurements, the proposed algorithm incorporates an innovative method based on the density-based spatial clustering of applications with noise (DBSCAN) algorithm. This not only significantly reduces the number of partitions to be considered but also accurately distinguishes measurements when targets are in group conformity and group division, thereby markedly improving the performance and precision of multi-target tracking. Simulation results demonstrate that the proposed algorithm is effective for tracking multiple group targets in scenarios with group conformity and group division.