Human action recognition in videos based on the Transferable Belief Model

This paper focuses on human behavior recognition where the main problem is to bridge the semantic gap between the analogue observations of the real world and the symbolic world of human interpretation. For that, a fusion architecture based on the Transferable Belief Model framework is proposed and applied to action recognition of an athlete in video sequences of athletics meeting with moving camera. Relevant features are extracted from videos, based on both the camera motion analysis and the tracking of particular points on the athlete’s silhouette. Some models of interpretation are used to link the numerical features to the symbols to be recognized, which are running, jumping and falling actions. A Temporal Belief Filter is then used to improve the robustness of action recognition. The proposed approach demonstrates good performance when tested on real videos of athletics sports videos (high jumps, pole vaults, triple jumps and long jumps) acquired by a moving camera and different view angles. The proposed system is also compared to Bayesian Networks.

M. RombautEmail:

---

Emmanuel Ramasso   is currently pursuing a PhD at GIPSA-lab, Department of Images and Signal located in Grenoble, France. He received both his BS degree in Electrical Engineering and Control Theory and his MS degree in Computer Science in 2004 from Ecole Polytechnique de Savoie (Annecy, France). His research interests include Sequential Data Analysis, Transferable Belief Model, Fusion, Image and Videos Analysis and Human Motion Analysis. Costas Panagiotakis   was born in Heraklion, Crete, Greece in 1979. He received the BS and the MS degrees in Computer Science from University of Crete in 2001 and 2003, respectively. Currently, he is a PhD candidate in Computer Science at University of Crete. His research interests include computer vision, image and video analysis, motion analysis and synthesis, computer graphics, computational geometry and signal processing. Denis Pellerin   received the Engineering degree in Electrical Engineering in 1984 and the PhD degree in 1988 from the Institut National des Sciences Appliquées, Lyon, France. He is currently a full Professor at the Université Joseph Fourier, Grenoble, France. His research interests include visual perception, motion analysis in image sequences, video analysis, and indexing. Michèle Rombaut   is currently a full Professor at the Université Joseph Fourier, Grenoble, France. Her research interests include Data Fusion, Sequential Data Analysis, High Level Interpretation, Image and Video Analysis.