Novel multi-step recursive sampling strategy for particle filtering in object tracking

In this paper, we propose a new sampling strategy for particle filtering in object tracking. Improper sampling can bring in heavy computation for achieving acceptable tracking results in particle filter. We propose a multi-step recursive sampling method to replace the direct importance sampling. This relies on the feedback from the resampling procedure. New particles are sampled recursively from the existing particles with high weights. After the iterations, particles become densely populated, and this approach contributes significantly to achieve accurate position estimation. Experimental results indicate that the method reduces computation substantially and it also preserves good tracking results.