Visual Tracking Applying Depth Spatiogram and Multi-feature Data

摘要

Object tracking, in general, is a promising technology that can be utilized in a wide variety ofapplications. It is a challenging problem and its difficulties in tracking objects may fail whenconfronted with challenging scenarios such as similar background color, occlusion,illumination variation, or background clutter. A number of ongoing challenges still remain andan improvement on accuracy can be obtained with additional processing of information. Hence,utilizing depth information can potentially be exploited to boost the performance of traditionalobject tracking algorithms. Therefore, a large trend in this paper is to integrate depth data withother features in tracking to improve the performance of tracking algorithm and disambiguateocclusions and overcome other challenges such as illumination artifacts. For this, we use thebasic idea of many trackers which consists of three main components of the reference model,i.e., object modeling, object detection and localization, and model updating. However, there aremajor improvements in our system. Our forth component, occlusion handling, utilizes the depthspatiogram of target and occluder to localize the target and occluder. The proposed researchdevelops an efficient and robust way to keep tracking the object throughout video sequences inthe presence of significant appearance variations and severe occlusions. The proposed methodis evaluated on the Princeton RGBD tracking dataset and the obtained results demonstrate theeffectiveness of the proposed method.