Real-time people detection and tracking using 3D depth estimation

Abstract

Detecting and tracking people in video have a wide variety of applications in computer vision. However the development of robust methodologies for these purposes are challenging due to complexity of the scenes. In this paper, we propose a real-time algorithm for people detection and tracking by tridimensional depth estimation in indoor environment. The approach is based on people detection and stereo processing techniques for 3D depth estimation of the analyzed scene. The prototype implementation showed satisfactory results for people detection and tracking within pre-recorded video and also in real-time captured video by a stereo camera. The performance on standard hardware and using open source software library achieved a frame rate of up to 7 fps.