Vehicle Tracking Incorporating Low-Rank Sparse into Particle Filter in Haze Scene
Wu Gang, Zeng Xiaoqin, Shoubao Su, Hong Lu · 2016
Aiming at difficulties for vehicle tracking in haze Scene, a new particle filter algorithm for tracking object vehicle is proposed in this paper. We incorporate low-rank sparse and adaptive dictionary learning template with the classical particle filter algorithm. We apply enhanced algorithm to track selected vehicle in urban traffic scene, demonstrate the performance of our method on the process of vehicle tracking in haze Scene. Our approach is different from conventional improved particle filter algorithm. Compared with related L1 and online MIL methods, experimental results show that the proposed low-rank sparse particle filter algorithm improves the synthesized efficiency of tracking system, the tracking experiments based on standard testing videos involving Dtneu_nebel demonstrate that tracking error is significantly reduced in the case of less time tracking.