In tracking closely-spaced targets, the kinematic information is inadequate to resolve the ambiguity of measure-to-track correspondence, which often leads to track coalescence or mistake association. In this paper, we propose a Fusion Tracking framework to a joint target tracking and recognition problem based on MTI and ESM. According to the time sequence of each sensor measurement, the corresponding point-track association module is carried out. Furthermore, the attribute fusion is performed using the Dempster-Shafer rule of combination and the reliability was also considered. Simulation results demonstrate that a joint framework outperforms the traditional single sensor strategy in terms of both recognition and tracking performances.
展开▼
