| Summary: | 碩士 === 國立中央大學 === 資訊工程研究所 === 97 === To touch and manipulate data on screens directly without any intermediary device provides more convenient and instinctive using experience to the user. Therefore, the touch screen has become a trend for the display technology in recent years, especially to multi-touch technique attracts more focus. In this work, we propose a vision-base touch screen architecture with dual cameras, touch detection and locating algorithms. For the fraud point problem, we adopt touch point tracking to solve it and achieve the dual touch ability. The ready-made liquid crystal display (LCD) can mount this dual cameras module to equip the dual touch ability, and the cost is cheaper and the volume is smaller than current multi-touch techniques.
Moreover, we define a set of touch gestures and map it to the controlling commands of mouse. This makes it can manipulate those application software designed for mouse. We also define a touch data transmission protocol between the touch detecting programs and touch software. By using above schemes, we realize the touch software which manipulated with dual touch point inputs.
Finally, we implement the dual touch system on the 19-inch and 42-inch displays, respectively. Then, to verify the feasibility of proposed method, we test the touch operations with various applications. Furthermore, we evaluate the usability of our system with a hit game experiment, and the result shows the proposed method provides a satisfying performance.
|
|---|
