This paper presents a hand gesture recognition system as a part of our virtual reality system called non-contact flight auxiliary (NCFAC) system. The system is developed using Bayesian neural network to translate hand gestures to corresponding commands and utilizes one hand gesture to prepare for collision detection. Cyberglove sensory glove and Flock of Birds motion tracker are applied to this system to extract hand features. The Bayesian neural network model is trained and tested with different sample groups. Experiment shows that our system is able to recognize 16 kinds of hand gestures with the accuracy of 95.6% and greater generalization capability. The system can also be extended and use other algorithms for future works.
Bayesian neural network approach to hand gesture recognition system
2014-08-01
172994 byte
Conference paper
Electronic Resource
English
A model-based hand gesture recognition system
British Library Online Contents | 2001
|Real-Time Hand Gesture Recognition for Robot Hand Interface
Springer Verlag | 2014
|Hand gesture recognition system based on laser ladar
SPIE | 2022
|Real-Time Hand Gesture Recognition for Robot Hand Interface
British Library Conference Proceedings | 2014
|