| Augmented reality technology has become a research hotspot in computer vision,computer graphics,human-computer interaction and other fields in recent years.Mechanical input devices such as keyboards and mice,including physical touches such as touch screens,require people to spend time learning and adapting,and are less friendly and humane to the augmented reality environment.Visual gesture interaction is one of the main ways of human-computer interaction,providing users with more interactive space,more interactive degrees of freedom and a more realistic interactive experience.Visual gesture interaction is one of the main ways to easily grasp and humanize human-computer interaction in the field of augmented reality.It provides users with more interactive space,more interactive degrees of freedom and a more realistic interactive experience.This paper presents a gesture recognition method based on SVM-HMM model,and then designs and implements a set of 3D dynamic gesture interaction system.The recognition method uses Leap Motion to capture hand information,such as palm position,finger position and fingertip speed.Combined with SVM static gesture classification ability and HMM dynamic gesture modeling capability,the recognition accuracy rate reaches 96.8%.This 3D dynamic gesture interactive system can recognize 5 kinds of natural dynamic gestures set in advance,and automatically determine the start and end of the gesture to achieve continuous real-time gesture interaction.Experiments show that the gesture interaction system is effective,each gesture recognition only takes 0.15 s,the overall performance exceeds the relevant research results,and the learning cost is very low,and ordinary users can quickly grasp the method.We used the Pygame engine to develop an augmented reality flappy bird game.Combined with this set of gesture interaction system,we can implement a variety of interaction logic between the user and the game,thus verifying the practicality of the gesture interaction method. |
PDF Full Text Request