Research On Static And Dynamic Gesture Design And Recognition Of Intelligen Robot

Since the first digital computer was born in 1946,computer technology has achieved unprecedented development both at home and abroad.It is not only a common computing tool,but also plays an extremely critical and important role in the new industrial revolution.With the development of computer technology,it has been widely used in various fields of the whole society.However,although computer technology has made great progress,it is still only a kind of advanced tool,which is controlled,controlled and managed by people.In the use of computers to complete the process of various tasks,there is a large part of the people and the computer to complete.Under this background,human-computer interaction(HCI)system came into being.With the progress of The Times,the continuous development of virtual reality technology and robot technology,the traditional human-computer interaction is gradually unable to meet the needs of the natural interaction between human and computer.This,to some extent,promotes gesture recognition based on machine vision.As a novel human-computer interaction technology,gesture recognition has received widespread attention and attention from scholars at home and abroad.Based on this,this study firstly USES the method of literature review to review the current research achievements on gesture recognition technology at home and abroad,and briefly summarizes the theories related to natural gesture recognition technology and sensor technology,in order to lay a theoretical foundation for the research of this paper.On the basis of related literature and theory research,based on gesture recognition problems such as complex and changeable environment,design a kind of complex background composite blank,approach,and rely on gestures background real-time switch image preprocessing mode,through the infrared,visible light,the compound different gesture recognition under the environment of the night and day,and then improve the static gesture recognition adaptability to the environment.In addition,based on the problems of inconsistent frame number and unfixed configuration of dynamic gestures,a multi-length memory network model(LSMT)method was designed,and Kinect skeleton information was combined to identify and classify the same dynamic gesture in different frame number and shape.Finally,based on the effective application of gesture recognition technology in command control of intelligent robot,an interactive system integrating intelligent service robot and gesture recognition is designed,and the validity of gesture recognition technology in command control of intelligent robot is verified by simulation.