Research On Target Tracking Algorithm By Biological Vision Mechanism

The research of target tracking method is not only an important application in the field of computer vision,but also the core technology of security,transportation and human-computer interaction.The difficult point of target tracking is that we should consider the appearance of deformation,illumination changes,fast motion,motion blur and background similar interference in an algorithm,and we also should make the tracking algorithm highly real-time.This paper mainly makes the following research on target tracking.1)For target tracking drift or lost in ambient illumination changes,such as color change of the object caused by the change of external light.An accurate and efficient object tracking algorithm is proposed from the perspective of tracking model,a cyclic sampling kernel correlation filter tracking algorithm is proposed.The experimental results show that the algorithm can effectively enhance the robustness of the features.The improved algorithm had good stability.2)In view of the complexity of the original tracking model and the lag of the tracking effect.A new algorithm named BIMT is proposed.The algorithm uses the biological visual memory mechanism incentive model to represent the construction of the target appearance model,simulates the human observation mechanism of the target information,and a structured SVM classifier is used to simulate the instantaneous memory working space to separate the target tracker from the background noise.The tracking results verify the practicability of the proposed tracking algorithm,prove that the biological vision based mechanism has broad application prospects.3)The algorithm is applied to the mobile robot based on BIMT.The system fuses a monocular camera with a control chip to obtain the real-time positions of target.The tracking results verify the practicability of the proposed tracking algorithm,prove that BIMT tracking algorithm has broad application prospects.Figure29;Table7;Reference 50...