Hand Tracking and Gesture Recognition for Human-Computer Interaction

摘要: The proposed work is part of a project that aims for the control videogame based on hand gesture recognition. This goal implies restriction real-time response and unconstrained environments. In this paper we present algorithm to track recognise gestures interacting with videogame. three main steps: segmentation, tracking recognition from features. For segmentation step use colour cue due characteristic values human skin, its invariant properties computational simplicity. To prevent errors add second step, tracking. Tracking performed assuming constant velocity model using pixel labeling approach. From process extract several features are fed finite state classifier which identifies configuration. can be classified into one four classes or different movement directions. Finally, system’s performance evaluation results show usability in video game environment.