基于序列霍夫变换的多模型滤波算法

Multi-model Filtering Algorithm Based on Sequential Hough Transform

针对目标运动复杂多变的滤波问题,在序列霍夫变换的基础上,研究了多模型滤波算法,提出了一种基于序列霍夫变换理论的多模型滤波算法。该算法将最常用的多种目标运动模型进行组合,构建为目标运动模型集,将目标模型的索引参数扩充至目标状态,使得目标状态可以根据模型转移概率在运动模型集中进行切换。在非线性条件下将其实现,仿真分析结果表明,该算法能够在目标新生信息及目标运动模型均未知的前提下完成对多目标的跟踪。

In order to solve the complex filtering problem of the target motion,a multi-model filtering algorithm based on sequential Hough transform theory is proposed.The most commonly used multiple target motion models are combined by the algorithm to construct a target motion model set,and the index parameters of the target model are expanded to the target state,so that the target state can be switched in the motion model set according to the model transition probability.It is implemented under nonlinear conditions.The simulation results show that the algorithm can complete the tracking of multiple targets under the premise that the new target information and the target motion model are unknown.