本篇論文主要利用基於梯度之瞳孔追蹤演算法來開發應用程式。目前瞳孔追蹤有許多相關的技術與裝置,本論文選用網路攝像頭來獲得輸入影像,並搭配合適的演算法來追蹤瞳孔的位置,其特色為易於取得、價格便宜、方便架設。追蹤影像人臉的瞳孔位置之後,再以此做為基礎進一步設計使用眼睛操控的應用程式。 作者在此篇論文研製了三個應用程式,第一個是利用眼睛選擇答案的問答應用;第二個是用眼睛控制螢幕畫面的閱讀應用;第三個是使用眼睛追蹤幾何物體的小遊戲。經過測試,當使用者在不移動頭部的前提下,這些應用都能運作得相當良好。作者希望這樣的研究能夠對於行動不便的人,特別如癱瘓、萎縮或是受傷嚴重等,手腳不便且無法說話的人們,能夠提供一些幫助。
In this thesis, the author uses gradient-based pupil tracking algorithm to develop applications. Currently, there are many related technologies and devices for pupil tracking. In this thesis, a webcam is used to obtain input face images, and a well-designed algorithm is used to track the pupil positions. The advantages are easy to obtain and set up, and it is very cheap. Based on the pupil positions of the image face, the author further design three eye-control applications. The first application is a question-and-answer application that uses the eye to select answers; the second one is a reading application that uses the eye to control the screen; and the third one is a small game that uses the eye to track geometric objects. After testing, these applications work quite well when the user does not move the head. The author hopes that such a research can provide some help to people with disability, especially those who are inconvenient and unable to move and talk.