Adaptive Anti-occlusion Moving Object Tracking of Intelligent Video Surveillance System

Fa-cang DU, Ying ZHANG, Meng-xin LI

Abstract


In recent years, moving target tracking technology has been widely used in video surveillance and other fields. But there are many factors that are not conducive to the system. Among them, the large area occlusion are the main factors leading to tracking failure. In addition, the interference of illumination and some similar colors is also an important factor affecting the tracking accuracy. Aiming at the disadvantage of moving target tracking, this paper proposes a dual discriminant adaptive tracking algorithm based on mean shift algorithm and Kalman filter algorithm. When the target is tracked with illumination, large area occlusion, long time occlusion, background interference and similar color interference, the algorithm can track the target well, with good accuracy and stability.

Keywords


Motion target tracking, Meanshift algorithm, Kalman filter, Bhattacharrya coefficient, Anti occlusion, Double discrimination


DOI
10.12783/dtcse/aicae2019/31440

Full Text:

PDF

Refbacks

  • There are currently no refbacks.