Based On The Key Technologies Of Visual Multi-touch System

Multi-touch (MT) technology is a part of the field of Human Computer Interaction (HCI). With the improving of the computer's processing power and storage capacity and the development of the human-computer interface, the gesture interaction, voice recognition and multi-touch become the new generation representation of the field of the HCI technology. The early study of MT is generally limited in laboratory research or some special applications, it too expensive to common users. With the improving of the computer processing power, the MT technology is no longer relying on expensive equipment. Some devices based on camera, sensor, capacitor and resistance have been published. Due to the advantages of low cost, easy to build, the vision-based MT becomes a hot spot research. There are many organizations and institutions in the world engage in the research, development, and advocacy of this technology. It is no longer a dream to have an own MT device, though the design principle and implement details remain mysterious to many users.Vision-based MT system is composed of hardware and software, the hardware is the basic of the system as interface and the software is responsible for image processing and application development. Firstly, this paper will do a compare about the existed vision-based MT technology and build a MT device based on FTIR technology. Secondly, there will be a detail study and analysis of some key technology of software include image processing, blobs detection and tracking and coordinate position based on the related OpenCV functions. Then, we will have an introduction about the research work of gesture recognition and application development. Finally, in order to meet the feature that this system not only can identify the multi touch points at the same time, but also can be operated by multiple users simultaneously, this paper have a discussion about how to realize a bigger size screen in the way of increasing the number of cameras and combine with image processing.The research proves that, the vision-based multi-touch technology remain have some problems about the accuracy and deficiencies and response affected by the equipments and environments, but it could complete the work of human-computer interaction smoothly. Combined with the development of application and the interface, the technology has enormous potential for application.