基于改进RCE和RBF神经网络的静态手势识别

针对手势识别的手区域分割、手势特征提取和手势分类的三个过程，提出了一种新的静态手势识别方法。改进了传统的RCE神经网络用于手区域的分割，具有更高的运行速度和更强的抗噪能力。依Freeman链码方向提取手的边缘到掌心的距离作为手势的特征向量。将上一步得到的手势特征向量作为RBF神经网络的输入，进行网络的训练和分类。实验验证了该方法的有效性和可行性，并用其实现了人和仿人机器人的剪刀石头布的猜拳游戏。

Static hand gesture recognition based on improved RCE neural network and RBF neural network

Hand gesture recognition usually includes hand image segmentation,features extraction,and hand gestures classification.In this paper,a new method is proposed to deal with the three phases of static hand gesture recognition.The traditional RCE neural network is improved and it is applied to hand image segmentation.The improved RCE neural network is proved to has running fast and strong ability of anti-noise.Freeman chain code is used to extract the distance from hand edge to the centre of the palm as feature vectors.Those feature vectors are used as the input of RBF neural network.Experiment results show this method is efficient and feasible.A scissors-paper-stone game between human and humanoid robot is developed by using this method.