Adaptive multiple cues integration for particle filter tracking

Visual object tracking in complex environments is a challenging task in smart surveillance fields. In order to enhance the robust tracking performance, many tracking algorithms based on multi cues integration have been proposed. However, how the multiple cues are fused during tracking is still an open issue. This paper integrates multiple cues into particle filtering framework for robust tracking in situations where no single cue is suitable. A novel quality function is introduced to evaluate the reliability of each cue. With the weights corresponding to the cue reliabilities, the combined likelihood is estimated as the weighted average of each cue. Experiments show that tracking with multiple weighted cues provides more reliable performance than single cue tracking.