Research On Static Gesture Recognition Technology Based On Vision

With the development of mixed reality technology and human-computer interaction technology,computer vision-based gesture recognition technology has become a research hotspot.However,the recognition rate and accuracy of existing methods need to be improved.In this paper,a new identification scheme is designed for the problems existing in current vision-based gesture recognition technology.The main research contents are as follows:1)Aiming at the problem that the existing scheme is greatly affected by the external environment and the hand itself,and the recognition effect is unstable,a gesture recognition scheme based on skin color threshold initialization and semi-contour technology is designed.2)In the skin color threshold initialization,in order to adapt to the detection of different skin colors,the HOG feature is used and the skin color threshold is initialized by the SVM classifier.In the pre-image processing,the segmentation and contour extraction of gestures are realized by the binary algorithm and the opening and closing operations in the morphological processing.In the fingertip calibration,a new algorithm based on semi-contour technology is designed to realize the detection and calibration of the finger end point.In the gesture classification,a new classification algorithm is designed for the fingertip peak,which realizes the recognition of 8 gestures.3)The experimental results show that the average recognition rate of the identification scheme designed in this paper reaches 95.25%,and the average time is 68.797 ms.Compared with the existing fixed threshold algorithm and full contour algorithm,the recognition rate can be increased by 7 times compared with the convolutional neural network algorithm.