Dynamic gesture recognition based on temporal modeling

In order to solve the problem of dynamic gesture recognition in traditional methods, a method of dynamic gesture recognition based on time series is proposed. The video is intercepted at several frames, the gesture information is segmented by modeling skin color, and a set of pictures corresponding to dynamic gestures are fused. The fused pictures contain time series information. Classify pictures based on convolutional neural network to realize the recognition of dynamic gestures. The experimental results show that this method has greatly improved compared with traditional recognition methods..