Improved RCE neural network and its application in human-robot interaction based on hand gesture recognition

RCE neural network has been applied in lots of areas because of its advantages, such as less training time, easy to study patterns, and never dropping into local minimization, especially in image segmentation. In this study, we conduct an adjustment algorithm for the traditional RCE neural network. The new RCE neural network runs faster and performs better in anti-noise than the traditional one. Then, in accordance with three stages of hand gesture recognition, we suggest a new method for static hand gesture recognition. Firstly, we apply the improved RCE neural network to hand image segmentation. Secondly, we use Freeman chain code to extract the distance from hand edge to the palm-center as feature vectors. Finally, we use those feature vectors as the input of RBF neural network and train the RBF neural network. Experiment results show this method is efficient and feasible. We develop a scissors-paper-stone game between human and humanoid robot using this method.