Object Recognition from Local Scale-Invariant Features
ICCV '99 Proceedings of the International Conference on Computer Vision-Volume 2 - Volume 2
Speeded-Up Robust Features (SURF)
Computer Vision and Image Understanding
BRIEF: binary robust independent elementary features
ECCV'10 Proceedings of the 11th European conference on Computer vision: Part IV
Global Trajectory Construction across Multi-cameras via Graph Matching
ICIG '11 Proceedings of the 2011 Sixth International Conference on Image and Graphics
Consistent labeling of tracked objects in multiple cameras with overlapping fields of view
IEEE Transactions on Pattern Analysis and Machine Intelligence
An association-based multi-target tracking method
Proceedings of the 2013 Research in Adaptive and Convergent Systems
| Hi-index | 0.00 |
In object tracking, methods based on a particle filter are widely used, but the technique alone often fails in various situations. Sometimes multi-camera systems using homography are tried to solve problems like occlusion. We propose an adaptive particle filter approach in a two-camera surveillance system. The proposed method is based on accurate homography, which is used to estimate the position of a tracked object in one camera view, using the position information in the other camera view. However, for each camera view, we divide particles into two groups - one group to follow the observed object in each camera view as usual, but the other group to follow the estimated position by homography. In this way, the particles placed at the estimated position could correct any mistakes or help solve other problems like partial/full occlusion and reappearance after temporary disappearance during tracking. Experiments show a good performance in tracking an object, even though the object is occluded or moves out of one camera view temporarily.