Abstract: This paper proposes a hand gesture recognition algorithm based on RGBD data. Firstly, the gesture segmentation algorithm which combines depth data with color data is used to segment the hand gesture area more precisely. Secondly, circularity,convex hull points and convex defect points, 7Hu moment features of the segmented static gestures are extracted. Lastly, SVM are used to recognize different static hand gesture. The experimental results show that the proposed method can effectively identify the five kinds of static gestures, and has strong adaptability to the environment.

Key words: gesture recognition, depth data, gesture segmentation, feature extraction, SVM