基于GPU的粒子滤波并行算法

GPU-based parallel particle filter algorithm

针对粒子滤波跟踪算法计算量较大,需要在跟踪准确性与计算效率之间做出妥协的问题,分析了粒子滤波算法的并行性,提出了基于图像处理单元(GPU)平台的粒子滤波并行算法.将传统粒子滤波算法与GPU有效结合起来,充分利用GPU并行运算的性能,加快粒子滤波算法的计算速度.对所提出算法的计算性能与普通串行算法进行了对比,实验结果表明该算法在不降低跟踪准确性的同时,平均每帧处理时间显著减少.

Particle filter algorithms for object tracking were widely studied because of its satisfying tracking effect and robustness. But high computational complexity of particle filter algorithms often results in the compromising between accurate tracking and computational efficiency, After analyzing the parallelism of particle filter algorithms, a parallel particle filter algorithm based on GPU was proposed. The computation performance of the proposed method and the traditional serial algorithm was contrasted and analyzed. Experiments show that the parallel algorithm presented can greatly improve the speed of the computation of each image frame.