Aiming at the problem of multi-camera tracking and object matching in overlapped regions, an algorithm of multi-camera moving object handover based on object mapping is proposed in this paper. At First, the video image of the ground in the corridor is mapped to a building floor plan chart, and then the center point of the bottom line of minimum external object rectangle is defined as the foothold of moving object. The position of foothold is used to represent the moving object in order to achieve the mapping of moving object in the building floor plan. When the moving object is tracked and matched in the overlapping field of multi-camera, it is judged whether it is the same moving object according to the mapping position in the building floor plan, the moving direction and the color histogram of moving object. So an accurate tracking object handover matching can be obtained. The continuous tracking of the same moving object can be achieved between different cameras. The experimental results show that the proposed algorithm can accurately obtain the real time position of moving object and realize the continuous tracking of the same moving object by the multi-camera.
[1]
Xiaogang Wang,et al.
Intelligent multi-camera video surveillance: A review
,
2013,
Pattern Recognit. Lett..
[2]
Zonghai Chen,et al.
Multi-camera handoff for person re-identification
,
2016,
Neurocomputing.
[3]
Yang Tao.
Fast Algorithm for Automatic Generating Field of View Line of Multiple Cameras
,
2007
.
[4]
Dariu Gavrila,et al.
A Comparative Study on Multi-person Tracking Using Overlapping Cameras
,
2013,
ICVS.