Extended object tracking and classification based on recursive joint decision and estimation
Wen Jia Cao, Jian Lan, X. Rong Li · International Conference on Information Fusion · 2013
Extended object tracking and classification (EOTC) involves both decision and estimation, where they affect each other. This is a joint decision and estimation (JDE) problem and good solutions require solving the two problems jointly. The recently proposed JDE and recursive JDE (RJDE) are preferable for solving EOTC problems. To describe the extended objects with different maneuverability, a new kinematic model specifying a constant-turn motion is proposed. This model fits well with the existing random-matrix-based EOT approach. Then the original point target RJDE is extended to EOTC with a multiple model approach. Further, two joint performance measures are provided to evaluate the performance of the proposed method. An illustrative example is elaborated, in which the RJDE approach is compared with traditional algorithms. To gain further insight into the RJDE property in EOTC, this paper analyzes the effect of parameters by comparing the performance of RJDE with E&D (optimal decision and optimal estimation, respectively) in different scenarios. Simulation results show that RJDE has the potential to beat E&D for EOTC.