State-of-the-art human detection methods focus on deep network architectures to achieve\nhigher recognition performance, at the expense of huge computation. However, computational\nefficiency and real-time performance are also important evaluation indicators. This paper presents\na fast real-time human detection and flow estimation method using depth images captured by a\ntop-view TOF camera. The proposed algorithm mainly consists of head detection based on local\npooling and searching, classification refinement based on human morphological features, and tracking\nassignment filter based on dynamic multi-dimensional feature. A depth image dataset record\nwith more than 10k entries and departure events with detailed human location annotations is\nestablished. Taking full advantage of the distance information implied in the depth image, we achieve\nhigh-accuracy human detection and people counting with accuracy of 97.73% and significantly reduce\nthe running time. Experiments demonstrate that our algorithm can run at 23.10 ms per frame on a\nCPU platform. In addition, the proposed robust approach is effective in complex situations such as\nfast walking, occlusion, crowded scenes, etc.
Loading....