Dynamic Gesture Recognition Based on the Trend of Key Points
Cai Mengmeng, Feng Zhiquan, Luan Min · 2017
This paper presents a new dynamic gesture recognition algorithm. Through the counter-clockwise obtain the edge coordinates of the hand, thereby obtaining theinformation like the location and distance, angle of the finger-root and fingertip. This information is used to identify commonly used dynamic gestures like grabbing, putting, moving, clockwise rotating, counterclockwise rotating and pushing forward, etc. This method is not affected of the moving speed of hand. And it can solve the problem of recognition error caused by poor segmentation, such as the influence of gesture internal voids and noise. A large number of experiments show that the algorithm proposed in this paper can identify nine kinds of commonly used dynamic gestures such as grasping, translating and clockwise rotation, counterclockwise rotation, and the average recognition rate is 97%.