Research On Key Technologies In Interactive Presentation System Based On Video Gesture Recognition

With the development of science and technology,human-machine interaction scenes have widely appeared in human daily life.Hand gestures,as the most natural way of expression for humans,have received widespread attention in the field of human-machine interaction.At present,the research directions of hand gesture recognition based on machine vision have been get fruitful results.But most current interactive presentation systems require dedicated hardware,and there are many deficiencies in the application of interactive projection presentations.This paper designs an interactive projection demonstration system based on video hand gesture recognition technology to solve the problems encountered in the current application of interactive demonstration systems in projection applications.This article first adds a video capture device to the projection demonstration system and designs an interactive projection demonstration system based on RGB video hand gesture recognition.An interactive region detection method based on the fusion of Sobel operator method and brightness difference method is proposed to accurately obtain the interactive region.The deep features of the image obtained by the convolutional neural network are used to recognize hand gestures,which improves the accuracy and robustness of the system's hand gesture detection and recognition performance.By performing Kalman filter tracking on the detection and recognition results of the specified hand gesture and calling the system interface,interactive functions such as key annotation and page turning are realized.This system improves the demonstrations experience of user,and has the advantages of simple hardware structure and easy installation.During the development process,it was found that the interactive demonstration system based on RGB video of hand gesture recognition,under the influence of complex projection images,lighting,etc.,the stability and accuracy of the system decreased.Considering that the depth image has good robustness to the above effects,an interactive demonstration system for hand gesture recognition based on Kinect V2 acquisition of depth video was designed.An interactive region detection method based on depth video feature clustering analysis is proposed.On the basis of ensuring the recognition accuracy,the depth video concise texture feature is used to simplify the structure of neural network and improved the speed of hand gesture recognition.After testing,the interactive presentation system based on depth video is more robust to a variety of projection presentation scenarios,recognition of hand gesture is stable,and it runs smoothly.