A Target Tracking Method Based on Dynamic Salient Features

Abstract

Given that dealing with blocking of traditional target tracking algorithm is not enough, a target tracking method fused into dynamic salient features is proposed by simulating human visual mechanism to ensure accuracy and efficiency. First, the salient features of the bottom layer image, such as color, intensity, orientation, and motion, are extracted. These features are considered feature vectors fused into target tracking algorithm. An improved target tracking algorithm is proposed because local regional histogram of target is influenced by the background pixels of the background region. Experiment results show that the proposed target tracking algorithm is more accurate than the traditional tracking algorithm in dealing with blocking, thereby meeting the needs of complex scenes.