Improvement And Research Of Gesture Recognition Algorithm Based On Leap Motion

As an important way of human-computer interaction,gesture recognition is very different from the input mode of other control devices.It has the characteristics of high recognition accuracy and simple input operation,and has huge advantages in machine equipment control and remote operation is the focus of current research.This paper introduces the development process of gesture interaction technology,and studies the current status of gesture interaction technology at home and abroad.Aiming at the traditional gesture recognition method,a gesture recognition scheme using Leap Motion is proposed,which solves the problems of insufficient light and occlusion of gestures in the previous gesture collection process,and improves the gesture recognition rate of the system.The traditional techniques for gesture recognition based on KNN algorithm mainly include KNN algorithm based on Euclidean distance and KNN algorithm based on Manhattan distance.Aiming at the problem that the traditional K-Nearest Neighbor Classifier(KNN)gesture recognition rate is generally low,this paper proposes an improved scheme.Firstly,the traditional KNN algorithm is improved by using the weight distribution principle,the KNN algorithm of weight allocation is compared with KNN based on European distance and KNN b ased on Manhattan distance respectively,the gesture recognition rate is improved,but the improvement effect is not obvious.Secondly,the KNN algorithm for weight distribution is fused with the Support Vector Machine(SVM),and the algorithm obtained after the fusion and improvement is named S-KNN algorithm.Finally,the S-KNN algorithm is compared with the KNN algorithm based on Euclidean distance and the KNN algorithm based on Manhattan distance.Finally,it is concluded that the gesture recognition rate of the S-KNN algorithm is significantly improved compared with the traditional KNN algorithm,which proves that Improved effectiveness.