Multiple Object Tracking Based on Adaptive Depth Segmentation

In this paper, we propose a multiple object tracking algorithm in three-dimensional (3D) domain based on a state of the art, adaptive range segmentation method. The performance of segmentation processes has an important impact on the achieved tracking results. Furthermore, segmentation methods which perform best on intensity images will not necessarily achieve promising results when applied on depth images from a time-of-flight sensor. Here, the employed unique segmentation promises a real-time tracking analysis, having a significantly high preprocessing efficiency. Our experiments confirm the robustness, as well as efficiency of the proposed approach.