Research On Gesture Recognition And Its Application In Teaching System

In recent years,with the development of computer technology,the interactive mode between people and machines is changing,and the use becomes more and more convenient.The interactive mode starts from the traditional keyboard based,mouse based touch control to the somatosensory interaction pattern based on action recognition.Human-computer interaction recognition technology has been applied to many areas of life.After years of change,now human-computer interaction is mainly to allow the machine to better adapt to people.With the continuous development of human-computer interaction natural language,gesture recognition in teaching has broad space for development.For visual recognition there are two kinds of cameras can be used,one is an ordinary network camera,one is a Kinect camera.Ordinary network camera because of the impact of light changes,poor recognition effect.This paper uses Kinect camera to improve recognition rate.In the experiment,the finger detection method of convex defect is proposed and applied to the static detection process.According to the characteristics of depth extraction the hand region produces two value image,and then use the Canny operator to extract two value image contour,according to the complete convex convex hull fingertip detection,through the convex defect detection to achieve fine finger points and its application in the identification of PPT pictures in.Dynamic recognition through Kinect access to the human skeleton,normalization of bone,joint trajectories through points as feature points of joint trajectory judgment,and then through the comparison of the three algorithms combined with the application of dynamic gesture recognition in the teaching process to determine the DTW algorithm as the final algorithm in this experiment,by limiting the search range further optimization of the DTW algorithm,so as to achieve the recognition of 4 kinds of motion gestures.Finally,the static gesture fingertip recognition and dynamic gesture trajectory recognition are applied in teaching.