Particle filter tracking based on fragment multi-cue integration

作者：Gu Lichuan; Liu Jianxiao; Wang Chengji; Cao Mengru

来源：International Journal of Applied Mathematics and Statistics, 2013, 51(24): 31-40.

• Adaptive cue integrations
• Adaptive modeling
• Applicable conditions
• Color cues
• Image fragments
• Multi cues
• Particle filter

A novel particle filter tracking algorithm is presented in this paper by using fragment multi-cue integration. The proposed tracker enhances the suggested FragTracker algorithm to employ an adaptive cue integration scheme. This is done by employing particle filter framework, each particle represents an image fragment described by edge and color cues, the weight of each particle is designed to be updated by adaptive multi-cue integration. Additional, the template of each fragment are updated during tracking by distinguishing occlusions and appearance changes, which is done by risk decision theory. The template is unchanged if the target is occluded;some fragments of template are updated if appearance change. In the experiments and comparisons, the applicable conditions of the proposed adaptive model are discussed, and its robustness and effectiveness are demonstrated.

• 出版日期2013