A robust and real-time multi-space occupancy extraction system exploiting privacy-preserving sensors

A novel, real-time and robust occupancy extraction system is going to be presented in this paper. The main contribution of the paper is to introduce a novel camera calibration algorithm to a multi-space area. The proposed calibration method is accurate, efficient and free of any error propagation among cameras' calibrations, without usage of overlapping areas. Also, the paper introduces a new virtual camera incorporated to the detection algorithm handling occupants' partial occlusion. The system is real time and utilizes only privacy-preserving sensors. The experimental results will illustrate the robustness and the efficiency of the system.

[1]  Pascal Fua,et al.  Multicamera People Tracking with a Probabilistic Occupancy Map , 2008, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[2]  Octavia I. Camps,et al.  Modeling Correspondences for Multi-Camera Tracking Using Nonlinear Manifold Learning and Target Dynamics , 2006, 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06).

[3]  Murtaza Taj,et al.  Distributed and decentralized multi-camera tracking: a survey , 2011 .

[4]  Isaac Cohen,et al.  Jeju Island , Korea TRACKING PEOPLE IN CROWDED SCENES ACROSS MULTIPLE CAMERAS , 2004 .

[5]  Alberto Cerpa,et al.  Occupancy based demand response HVAC control strategy , 2010, BuildSys '10.

[6]  Leonidas J. Guibas,et al.  Counting people in crowds with a real-time network of simple image sensors , 2003, Proceedings Ninth IEEE International Conference on Computer Vision.