experiments results in paper "Never Worry about Occlusions: A Regional Color Histogram Based Occlusion Estimating Agency to Deal with Occlusions in Tracking"

Visual tracking methods have achieved a successful development in recent years. Especially the Discriminative Correlation Filter (DCF) based methods have significantly advanced the state-of-the-art in tracking. The advancement in DCF tracking performance is predominantly attributed to powerful features and sophisticated online learning formulations. However, it would come to some troubles if the tracker learns the samples indiscriminately. Especially when the target is totally occluded, the tracker will continuously learn the wrong information and likely lose the target in the following frames. In this work, we proposed a Regional Color Histogram Based Occlusion Estimating Agency (RCHBOEA), with the aim of avoiding wrong training of filter and improving robustness when the totally occlusions occur. Our method could be easily combined with any trackers which need online update. We performed experiments on a number of characteristic sequences which include severe occlusions. The results showed that the tracker combined with our agency achieved a remarkable performance.