An Anti-occlusion Object Tracking Algorithm Using KCF and ORB Feature Detector

To address the problem that Kernelized correlation filter (KCF) has poor tracking effect when the target is occluded, this paper proposes an anti-occlusion tracking algorithm based on Oriented FAST and Rotated BRIEF (ORB) and KCF. In the initial stage of tracking, the target features are extracted and feature templates are generated by ORB, and then the target is tracked by KCF. If the target is occluded, the training and updating of the filter are stopped, the target features in the current tracking frame are extracted by ORB, and then the transformation relationship between them is obtained by matching the features with the template features. The exact position of the target in the current frame is then obtained. The experimental results show that the precision and success rate of our proposed algorithm improve by 18.85% and 19.23%, respectively, compared with the KCF algorithm, which effectively improves the tracking effect of KCF in the case of target occlusion.