Research On Gesture Recognition Technology For Intelligent Service Robot

In recent years,with the rapid development of science and technology,intelligent service robots have come into being.Through continuous accumulation and breakthrough,a large number of intelligent service robot products and equipments have gradually integrated into the work and life of the public.Human-robot interaction is a key technology in intelligent robots,and gesture recognition is an important method to achieve friendly,natural and humane communication between human and robots.Gesture recognition interprets interactive meanings contained in hand and arm postures,or movements,so as to achieve a natural communication between human and robots.This article focuses on the gesture recognition technology of the intelligent service robot and its application in human-computer interaction,and carries out the following researches:First of all,for the complicated and changeable environment of gesture recognition.A compound processing method to deal with both blank background and complex background is designed.The image preprocessing mode can be switched in real time based on the gesture background.Besides,the combination of visible light and infrared light achieves the compatibility of gesture recognition both during the day and night time,thus enhancing the adaptability of static gesture recognition system to the environment.Secondly,for the problem of inconsistent configuration of dynamic gestures and inconsistent number of motion timing frames.The introduction of Dynamic gesture recognition method based on multiple LSTM network models,combined with Kinect skeleton information,enables to identify the classification and recognition of dynamic gestures with different forms and different timing frames.Finally,a set of gesture recognition and intelligent service robot interaction system was designed.By simulation,it verifies the feasibility and effectiveness of the gesture recognition technology when controlling the intelligent service robot.