摘要 在這篇論文中,我們主要是提出利用網路攝影機、簡單的材料和影像處理等來達成觸控板的功能。目前坊間的觸控板常見的觸控板依照感應原理大致上可分為電阻式、電容式和音波式等等,隨著需求不同而選用不同的觸控板。而本研究將提出不同於前述形式的觸控板。 本研究將利用網路攝影機搭配影像處理技術並在實驗中結合WINDOWS中的游標,來實現出另一種形式的觸控板,因此在本研究將有以下兩個重點。第一 ,滑鼠游標之定位,在這裡我們將提出如何利用灰階影像、影像相減、影像分割等影像處理技術來實現滑鼠游標之定位。其中,隨著使用環境光線的不同還有使用者習慣的不同,在做滑鼠游標定位時我們將會面臨雜訊的問題,在這裡我們提出利用物件連通法來降低雜訊的干擾,使得滑鼠定位更加的精準。第二,我們將討論如何利用影像處理來定義左鍵和右鍵。 在這篇論文中,將具有以下的貢獻: 一、我們只需使用一個基本的網路攝影機和一些簡單的材料就可 以架設出一個成本較為低廉且功能不遜於一般坊間的觸控 板。 二、利用物件連通法來降低雜訊的干擾。 三、我們利用網路攝影機搭配影像處理將使得我們的觸控板具有 滑鼠左鍵和右鍵點選的功能。 四、除了實現觸控板基本功能外,這也讓我們利用網路攝影機所 建構出來的觸控板能更加適合使用在各種的環境和使用者。
Abstract In this thesis, our point is the achievement of touch panel by using web camera, simple materials and image processing techniques. The touch panel is possible to divide into the resistance type, capacitive type and surface acoustic wave type according to the induction principle. By different needs and choosing different touch panel. This research will address format of touch panel which is different from prior type. This thesis will implement another format of touch panel by using web camera, image processing techniques and combining with windows mouse cursor in an experiment. Therefore, we have two emphases in this research. First, the mouse cursor positioning will be considered. We will achieve mouse cursor position by using gray image, image subtraction, image segmentation, and other image processing techniques. In addition, the noise reduction will be taken into consideration. While the difference of environmental light and user habits is encountered, we will face the problem of noise in mouse cursor position. Here we will reduce noise by using connected component labeling to make mouse cursor position be more accurate. Second, we will discuss how to use image process techniques to define the left button and the right button. By these emphases, the touch panel with web camera will be established. There are four contributions in this thesis. First, we implement a cheaper touch panel by using a web camera, simple materials and image processing techniques. Second, the noise can be reduced by using connected component labeling. Third, we combine web camera with image processing to make touch panel with capability of left button and right button to execute. Finally, the touch panel proposed by the authors will be more suitable to any kinds of environments and users.