Title :
Towards real-time markerless human motion capture from ambiance cameras using an hybrid particle filter
Author :
Fontmarty, Mathias ; Lerasle, Frédéric ; Danès, Patrick
Author_Institution :
LAAS-CNRS, Univ. of Toulouse, Toulouse
Abstract :
In this article we present a two ambiance video camera system dedicated to markerless human motion capture. We introduce a new particle filter algorithm which entails an importance function enabling auto-(re)initialisation, and takes account of the global curvature of the likelihood so as to guide the search along poorly observable directions of the state space. The system robustness is improved by fusing different visual cues. Performances are nearly real time.
Keywords :
image fusion; image motion analysis; importance sampling; object detection; particle filtering (numerical methods); tracking filters; video cameras; ambiance video camera system; hybrid particle filter; importance sampling; real-time markerless human motion capture; state space; visual cues fusion; visual tracking system; Annealing; Cameras; Filtering; Humans; Monte Carlo methods; Particle filters; Real time systems; Robustness; Sampling methods; State-space methods; data fusion; human motion; particle filtering; visual tracking;
Conference_Titel :
Image Processing, 2008. ICIP 2008. 15th IEEE International Conference on
Conference_Location :
San Diego, CA
Print_ISBN :
978-1-4244-1765-0
Electronic_ISBN :
1522-4880
DOI :
10.1109/ICIP.2008.4711853
