关键词: Canny算法, 手势识别, 小波矩, 文字输入系统

Abstract: Vision-based gesture recognition is a key technique to achieve a new generation of human-computer interaction. As few text input search system by gesture recognition is developed, based on the existing gesture recognition techniques, this paper uses the gestures which are corresponding to the Chinese letters and numbers as input gesture and uses Microsoft kinect to obtain depth image to conduct hand gesture segmentation. The edge of the gesture is extracted by Canny algorithm, and then the feature is extracted based on wavelet moment. The gesture letters are obtained. It achieves the text input system based on gesture recognition. Experiments show that the system is able to achieve Chinese characters accurately and effectively.

Key words: Canny algorithm, gesture recognition, moment compare, text input system