【摘要】 基于蒙特卡洛模拟的粒子滤波算法被广泛地应用于目标追踪领域。传统的粒子滤波算法在其追踪过程中所使用的粒子数通常是固定不变的,而在实际应用中,这会使算法缺乏高效性。针对这个问题,提出了一种自适应性粒子滤波器,它可以根据实际调整算法运行过程中使用的粒子数目,以使算法在保持对目标进行有效追踪的同时节省计算资源。仿真结果显示了算法的高效性。更多还原

【Abstract】 As an efficient algorithms particle filter has been widely used in the area of object tracking.Normally number of particles being used in the algorithm is given as a constant which would somewhat make the algorithm unstable and inaccurate.As that an adaptive particle filter is proposed,this motion model and particle number can change adaptively according to the object’s moving state.And the simulation shows that our algorithm is very effective when used in the tracking program.更多还原