Robust human tracking based on multi-cue integration and mean-shift
Pattern Recognition Letters
Visual tracking algorithm based on CAMSHIFT and multi-cue Fusion for human motion analysis
SMC'09 Proceedings of the 2009 IEEE international conference on Systems, Man and Cybernetics
Human tracking: a state-of-art survey
KES'10 Proceedings of the 14th international conference on Knowledge-based and intelligent information and engineering systems: Part II
Handling sequential observations in intelligent surveillance
SUM'11 Proceedings of the 5th international conference on Scalable uncertainty management
| Hi-index | 0.00 |
Visual tracking has been an active research field of computer vision. However, robust tracking is still far from satisfactory under conditions of various background clutter, poses and occlusion in the real world. To increase reliability, this paper presents a novel Dynamic Bayesian Networks (DBNs) approach to multi-cue based visual tracking. The method first extracts multi-cue observations such as skin color, ellipse shape, face detection, and then integrates them with hidden motion states in a compact DBN model. By using particle-based inference with multiple cues, our method works well even in background clutter without the need to resort to simplified linear and Gaussian assumptions. The experimental results are compared against the widely used CONDENSATION and KF approaches. Our better tracking results along with ease of fusing new cues in the DBN framework suggest that this technique is a fruitful basis to build top performing visual tracking systems.