结合归一化中心矩与DAGSVM的机器人交互手势识别

ROBOT INTERACTIVE HAND GESTURE RECOGNITION COMBINING NORMALISED CENTRAL MOMENTS AND DAGSVM

将归一化中心矩和DAGSVM相结合应用于机器人人机交互中的手势识别。归一化中心矩作为手势特征具有平移和比例不变性,同时对方向变化比较敏感,这正是机器人识别不同指向的手势时特征所需具有的特性,然后,将这个手势特征向量输入到DAGSVM分类器进行分类识别。实验结果表明,该方法能够有效地识别手势,且用于控制机器人运动效果良好。

We combine the normalised central moments and the DAGSVM （directed acyclic graph support vector machine） together to apply them to hand gesture recognition in robot-human interaction. The normalised central moments is selected as the feature of hand gestures, which is robust to translation and scale, at the same time it is also sensitive to direction, and these are the necessary characters for robot to recognise different pointing hand gestures. Afterwards, we input the gesture feature vectors to DAGSVM classifier for classifying and recognising. Experimental results show that this method can effectively recognise the hand gestures, and has good effect when to be used to control the robot motion.