Implementation Of Multi-Camera Collaborative Tracking System With Application Of Multiple View Geometry

The goal of object tracking is to imitate the motion sensibility of human vision, empower the machine with the ability of perceiving the target motion in the image sequence, and provide an important data source for visual analysis and understanding. Along with the rapid development of surveillance system, more and more cameras being used, multi-camera collaborative object tracking has become one of the most important issues in object tracking.Multi-camera collaborative object tracking is a computer vision technology also known as relay tracking which is to track some interesting object appeared in several fields of view captured by multiple cameras. Its robustness is boosted by expanding observation area and observation angle.Relation among cameras is vitally needed in multi-camera collaboration. In order to acquiring these relations, epipolar geometry and homograpyh principle must be applied to calibrate cameras and compute correspondence between imaging plane with another one as well as with the real world. Therefore, the application of multiple view geometry provides theoretical basis and technical support to realizing a multi-camera collaborative object tracking system.In this paper, the research is focused on multi-camera collaborative object tracking. This paper mainly includes three parts:Firstly, applying multiple view geometry to calibrate cameras and solving correspondence between imaging plane with another one as well as with the real world, computing parameters of camera model such as intrinsic matrix, distortion coefficient, homography matrix, etc.Secondly, proposing a multi-camera object collaborative tracking algorithm based on position-sensed discriminative appearance model, which is an algorithm gives consideration to both accuracy and efficiency.Thirdly, designing and realizing a multi-camera collaborative object tracking system in a scene of large parking lots, evaluating the effectiveness of the proposed multi-camera object collaborative tracking algorithm based on position-sensed discriminative appearance model.