Research On The Technology Of Interactive Touch System Based On Image Processing

At present, the human-computer interaction system began to appear widely in human’s daily life, and touch technology became one of the most popular research direction in human computer interaction.Touch technology greatly to subvert the original human machine operation, the communication between people and machines more smooth, also in a great extent improves of the interactive system is efficient, intelligent and friendly.Existing touch technology is quite extensive, such as resistive touch screen, capacitive touch screen, infrared touch screen, surface acoustic wave touch screen and optical touch screen. The most on the market currently is capacitive touch screen and the resistive touch screen. Optical touch screen has the advantages of low cost, high precision and convenient assembly, and has great potential for future development in the field of large touch screen.The object of this paper is an interactive touch system based on image processing. The aim is to build a stable, fast and accurate access to multi-touch information and real-time tracking and identification of user gesture human-computer interaction optical touch system. The main work of this paper is the design and implementation of the PC algorithm and the algorithm of multi contact detection and localization algorithm and gesture recognition.The system is mainly divided into three major modules: multi contact detection and positioning module, multi contact digital image processing module and multi touch gesture tracking and recognition module.Multi-touch detection positioning module mainly uses infrared optical touch technology is the basic principle, according to the characteristic of the system multi contact detection the four optical sensors to ensure the whole range of touch screen can be infrared coverage and when contact occurs can be optical sensor to detect, do the technical basis for the coordinate position of the contact. In this paper, a series of contact waveforms are processed to extract more accurate contact information. The coordinate information of the contact points is calculated by using the grouping method, and the more accurate contact coordinate values are calculated according to the error repair formula proposed in this paper.Multi-touch digital image processing module involves a large number of digital image processing technology, including video frame extraction, background difference, frame difference image, median filtering, image sharpening, morphological processing, edge detection, feature extraction and other related technical principles for processing the collected video image provides a powerful technology support. After the complex processing of the video frame image, the contact edge detection of the image and the successful extraction of the main goal of this system, to achieve the multi contact edge extraction algorithm in this paper.Multi-touch gesture track recognition module is one of the key content, according to the object of the system for moving contact this important feature, the system using motion tracking and recognition algorithm of video acquisition to image tracking and recognition, the record of the object motion trajectory tracking, through a series of coordinate calculation, so as to identify the user gestures and complete the corresponding instruction, the ultimate realization of the system of multi touch gesture tracking and recognition algorithm.At the end of this paper, the whole system is connected with the related performance test. The experiment results show that the whole system has completed the positioning detection and gesture recognition. Achieve multi-touch positioning response speed is less than 85 ms and gesture recognition response time maintain 150 ms within the whole system in accurate rate higher than 98%, the error is less than 3mm, stable work target, the system with more recognition, low cost, large screen, build is simple, fast response and other advantages.