Research On Vision-based Gesture Recognition Algorithms And Applications

Human-computer interaction technology has always been the research focus under the rapid development of computer technology in today's society,and it has a wide range of applications in many fields.Traditional human-computer interaction technology is mainly limited by hardware equipment,which makes human-computer interaction technology based on visual gesture recognition technology emerge as the times require.While the human-computer interaction method becomes more friendly and natural,it is not due to the greater limitations of hardware equipment?It is widely used in many fields and has become one of the mainstream research fields in the current human-computer interaction technology.Under the above background,there are five parts of traditional gesture recognition technology: gesture acquisition,image preprocessing,gesture detection,feature extraction,gesture recognition,and this article draws the corresponding advantages of image preprocessing methods in traditional gesture recognition technology,then designs a deep learning convolutional neural network model to train and judge gesture images.The related algorithms of gesture recognition image processing are studied and compared,and the skin color of the hand is extracted based on HSV color space and YCb Cr color space respectively.Aiming at the advantages and disadvantages of the two methods,a skin color extraction fusion algorithm is proposed.Combining the advantages of the two methods,the skin color extraction of the hands achieves a good processing effect.Aiming at problems such as insufficient accuracy of the Le Net-5 classic network model,it was improved and the gesture image was trained to achieve a good training and judgment effect.On the basis of the above model as the host computer,an application of gesture recognition technology is carried out,and the information obtained by judging the gesture is used as an execution instruction,and it is sent to the lower computer through wireless communication technology to control the device.In the image processing stage of dynamic gestures,aiming at the spatiotemporal difference between the gesture part and the background image a dynamic gesture skin color extraction combined algorithm based on background difference method and HSV color space is proposed to segment the gestures from the complex background relatively completely.It uses the above model to judge the gestures,and then use Wireless communication technology disseminates data and executes related instructions.This article conducts related research on the vision-based gesture recognition technology and application,optimizes the image processing algorithm,and improves the recognition effect.Based on this,the gesture recognition system is applied to the smart lamp,which has strong practicability and research value.