Recognition And Semantic Representation Techniques Of Gesture Behavior

With the development of science and technology,the application of human-computer interaction technology has become more and more widespread,and the "people-oriented" interaction concept has gradually attracted attention.Gesture interaction in human-computer interaction is one of the most mainstream interaction methods.Gesture interaction through vision is widely used in many fields such as automotive,medical and somatosensory games.At the same time,the rapid development of artificial intelligence has taken vision-based gesture recognition technology to a new level.Therefore,this paper integrates technologies in many fields such as artificial intelligence,cognitive science,and virtual reality to realize gesture recognition based on ordinary two-dimensional camera images,and use natural gestures to control drone flight in virtual systems.In order to realize gesture recognition under the two-dimensional camera,this paper uses deep learning technology to extract human bone joint points.And this paper designs a Pose-Hand model,which can extract all joint points and joint hands of the human body,which lays a good foundation for gesture recognition.In the design of gesture recognition algorithms,two methods based on positional relationship and deep learning are used for gesture recognition,which can effectively and quickly recognize simple gestures and complex gestures.At the same time,in order to reflect the "people-oriented" concept,this paper also studies the gesture semantic representation technology of gesture-controlled drones.Through the behavior experiments of gesture-controlled drones,a large amount of experimental data is collected to establish a semantic representation model of gesture-controlled drones.Finally,in order to verify the accuracy of the gesture recognition model and the gesture semantic representation model,a virtual simulation system for gesture-controlled drones was also built,and the effect of gesture-controlled drones was demonstrated through the virtual simulation system.