Design And Implementation Of Advertising Media Screen Interactive Software Based On TOF Sensor

With the rapid development of computer technology,human-computer interaction is developing towards the natural user interface gradually.Natural user interface means that users interact with computers in the most basic and common ways,such as emoticons,gestures,languages and so on.The application of human gesture has been widely concerned by researchers and is developing rapidly.Early computer vision systems were based on color images,which were harsh to lighting conditions,inefficient and ineffective.Depth image is a rapidly developing image representation method in recent years.TOF is a kind of technology that depth sensor gets distance of the object using the time lag or phase difference between reflected near-infrared light and emissive near-infrared light.The traditional advertising media screen can only complete the simple function of disseminating information,and cannot interact with users in two directions.It can only accept information in one direction passively,and it lacks interactivity.In order to meet the interactive requirements of various scenarios,this thesis designs an advertising media screen interaction system based on the EPC660 TOF sensor developed by ESPROS,including gesture interaction module,remote interaction module and touch screen interaction module.The gesture interaction module can identify five kinds of static gestures,and send gesture command to control advertising media screen.The remote interaction module can control the advertising media screen remotely,and the touch interaction module can control the advertising media screen through the touch screen.This thesis first introduces the man-machine interaction technology and TOF sensor developing situation briefly,analyzes the function of the advertising media screen requirements.Then,give the overall design of software and hardware according to the functional requirements,introduces its technology background in view of each module of the software.Then,a gesture recognition method based on depth image is studied.The number of fingers and the center distance feature are fused as feature vectors,template matching method based on weighted value is used to recognize hand gestures.Finally,the design and implementation of interactive system are introduced,and the system functions are tested and summarized.