An Airport Surface Target Tracking Method Based on Improved Meanshif

Abstract

To effectively solve the moving target losing problem of the change with background and tracking fast moving target in real-time airport surface video surveillance system, an automatic target tracking method based on improved Meanshift is proposed. Firstly, according to the analysis and research of the traditional Meanshift algorithm, motion vector is introduced to modify the candidate template center to lock target. In addition, color feature and edge feature is introduced to decrease the influence of changing surroundings or similar color. Finally, moving target is tracked accurately in real-time video surveillance system. The experiment on multiple targets (i.e. pedestrians and aircraft) in airport surface different surveillance scene shows that the proposed method can not only track moving target continuously and accurately, but also be applied in the foggy low visibility conditions.