Anti-Occlusion Tracking Method Based on Kernel Correlation Filter

Abstract

In the real scene, the target itself and the background will be unpredictable changes, such as the occlusion and other factors will bring greater challenges to the object tracking.In this paper, we proposed a method that handle the occlusion based on Kernel Correlation Filters (KCFFR, KCF with forward response).An anti-occlusion model is established based on the distribution feature of the forward classifier in response to the maximum value of the forward classifier.The target is searched by the block region helix search method after the target is occluded, and the response of the sliding box is calculated in the target search process to determine whether the target is found.The algorithm is tested on the OTB (Object Tracking Benchmark) sequence and compared with the four tracking algorithms.The results show the effectiveness and accuracy of the KCFFR algorithm.