Building the method for multi-target tracking based on the combination of PHD filter and JPDA filter using particle filter in 3D mixed coordinate system

  • Hung Nguyen Institute of Electronics, Academy of Military Science and Technology
  • Thang Pham Institute of Electronics, Academy of Military Science and Technology
  • Lam Tang Military Technical Academy
  • Tien Nguyen Military Technical Academy
  • Hưng Phạm Military Technical Academy
Keywords: Multi-target tracking; Mixed coordinates; PHD combined JPDA; Particle filter; Non-Gaussian; Constant velocity model.

Abstract

Changing the target number, non-linear measurement models and non-Gaussian noise faces a challenge to multi-target tracking problems which are factors affecting the accuracy, execution time and deciding the success of the method as well. In this paper, the authors present a method to solve these problems. Wherein, the motion of targets is represented in the mixed coordinate system 3D base on combining PHD (Probability Hypothesis Density) and JPDA (Joint Probability Data Association). This method can track multiple targets in the most general case, that is to change the target number, system model and measurement model which is non-linear as the noise is non-Gaussian. The result of this work can be applied to the real-time response system when the targets are moving in close distances with rapid maneuvering.

điểm /   đánh giá
Published
2023-08-25
Issue
Vol. 89 (2023)
Section
Overview