Detection of Defined Human Poses for Video Surveillance

This paper presents a system for real-time detection of defined human poses (i.e. raising of hands) in surveillance video. A single (non-calibrated) video camera is used to record data in an indoor environment. There are two main steps in our proposed system, the extraction of human silhouettes in video data and pose classification. Silhouette extraction is refined by paying attention to the removal of shadow artefacts close to occlusion borders. For pose classification, we combined, adjusted, and implemented two existing methods (star skeleton calculation and its evaluation). We demonstrate that the proposed two-step technique is solving the given task for a large percentage of input data when recording an individual person only.

[1]  Hironobu Fujiyoshi,et al.  Real-time human motion analysis by image skeletonization , 1998, Proceedings Fourth IEEE Workshop on Applications of Computer Vision. WACV'98 (Cat. No.98EX201).

[2]  Larry S. Davis,et al.  W4: Real-Time Surveillance of People and Their Activities , 2000, IEEE Trans. Pattern Anal. Mach. Intell..

[3]  Nigel J. B. McFarlane,et al.  Segmentation and tracking of piglets in images , 1995, Machine Vision and Applications.

[4]  Ferran Marqués,et al.  Silhouette-based probabilistic 2D human motion estimation for real-time applications , 2005, IEEE International Conference on Image Processing 2005.

[5]  Hong Liu,et al.  Detection of hand-raising gestures based on body silhouette analysis , 2009, 2008 IEEE International Conference on Robotics and Biomimetics.

[6]  Wilfried Philips,et al.  Detection of a hand-raising gesture by locating the arm , 2011, 2011 IEEE International Conference on Robotics and Biomimetics.

[7]  Chaur-Heh Hsieh,et al.  Human action recognition using silhouette histogram , 2011 .

[8]  Vivek Kumar,et al.  Human Activity Tracking using Star Skeleton and Activity Recognition using HMM ’ s and Neural Network , 2014 .

[9]  Bok-Suk Shin,et al.  Accurate Silhouette Extraction of a Person in Video Data by Shadow Evaluation , 2014 .

[10]  Reinhard Klette,et al.  Concise Computer Vision , 2014, Undergraduate Topics in Computer Science.