Moving Target Location Method Based On Euclidean Distance And Camshift Algorithm

Mingzhu Liu,Huixian Chen,Zhiqiang Qiu,Xiangxiang Ren

DOI: https://doi.org/10.1109/IMCCC.2018.00123

2018-01-01

Abstract:Moving target location and tracking is a very complicated problem, especially, the occlusion problem gradually becomes the limitation of the practical tracking algorithm. Because of the moving targets in video monitoring can be occluded by the obstacles in the moving process, it can reduce target positioning accuracy, and even lead to the target track losing. To resolve the problems, a new location algorithm has been proposed in this paper. The preprocessing of video image is realized through the basic process of Gauss foreground detection, Kalman filter, expansion and occlusion judgment. And then, the method based on the Kalman prediction and the adaptive threshold value which determined by the Euclidean distance of the target motion trajectory, and the target centroids have been searched by Camshift algorithm. The test results show that the method of using the Euclidean distance of the target trajectory as the prediction threshold can not only effectively locate the moving target, but also resist to the adverse effects of the occlusion object, and reduce the loss rate of tracking.