|
Touch technology as a nascent human-computer interaction, in the industrial, recreational and personal digital equipment being more widely used. Due to its user-friendly mode of operation, a good user experience, touch technology either for professional firms or ordinary users, have considerable appeal. However, in the past are just simple single contact operation, just to achieve a normal mouse emulation, greatly weakened its charm. In this thesis this scenario, the multi-touch technology in the fingers positioned to do a detailed discussion, study the problems which may be encountered, gives a preliminary application examples. Thesis focuses on multiple touch finger information collection and analysis process, involving optical imaging, mapping and camera intersection plane positioning, and information on the application of the touch of a preliminary exploration. In general, multi-touch technology includes three aspects: First, the positioning of multiple fingers point, two points are more than one finger tracking process, three are various definitions gesture recognition. Multi-point positioning are two other aspects of the foundation, but also to deal directly with the level of user input forms. Currently, there are many ways to implement the positioning of the finger, including capacitive sensing, sensing resistors, infrared, ultrasound, and optical based imaging techniques. In this paper, based on optical imaging positioning technology, an ordinary CMOS camera as the image acquisition equipment, learn computer vision measurement methods, the use of multiple cameras triangulation principle, to achieve the surface of the screen a certain precision positioning and simple demonstration of a multi- touch applications, with a strong novelty and practicality. The focus of this thesis and the difficulty lies in the images collected distortion correction, rejection and false finger point finger point tracking, and build the entire system. Image processing non-ideal lens distortion correction image obtained, corrected image point position of the camera to match the model definition. False finger point in occlusion culling process and multi-finger point finger circumstances exist false point. Finger point finger point tracking is obtained under the premise opponents pointing mark recording the movement in order to achieve more complex gesture recognition. Multi-touch gesture recognition is the ultimate goal of interactive input, gesture recognition can be used for more advanced applications. These features and the entire system in support of each module that implements a preliminary application.
|