部份身障者〈如漸凍人(肌萎縮性側索硬化症)或患有大腦性麻痺之病人等〉的肢體及臉部肌肉無法自主控制，極度影響他們本身及其家人的生活。若身障者只剩下眼睛仍能自主活動，透過使用追瞳系統，不僅可提升其獨立自主的能力，並可加強與人溝通的能力。現今已有數種追瞳系統問市，各有其優缺點。其中有些系統需使用者戴上頭盔，造成使用上負擔及不便；而複雜的操作流程也會增加使用者的挫折感，導致再度使用的意願不高；甚至有些系統是用高價來換取精準度，使得一般使用者無力負擔。基於以上之考量，本篇論文提出一套低價位追瞳系統，利用這套系統可以找到瞳孔中心座標，並估算出人眼凝視銀幕的相對位置，藉由對應後的座標控制滑鼠的移動。這個系統首先將CCD camera所擷取到的影像二值化，再利用投影機制找出瞳孔粗略位置，並運用邊緣偵測的方法得到瞳孔邊緣，接著使用一個簡單的群聚演算法求得瞳孔中心點。最後我們將所求得的瞳孔座標做網格點的轉換，找出人眼凝視銀幕的位置，並將其應用於人機介面的操作上。從實驗結果可以得知此系統可以快速且精確地執行追蹤瞳孔的工作，再搭配所設計之應用程式後，便可以有效地幫助身障者，提升其獨立自主的能力。 For some disabled, an extreme disability such as severe cerebral palsy or amyotrophic lateral sclerosis (ALS) deprives them of the use of their limbs and facial muscles. If eye motion is unaffected, the disabled could rely on an eye-tracking system to attain some degrees of independence and communicate with somebody more easily. Nowadays there are several different eye-tracking systems. Each of them has it’s own advantages and disadvantages. Some systems ask users to put on a helmet on the user’s head. Wearing a helmet will probably result in an uncomfortable feeling and increase the operational load. The other systems involve a complex operation procedure that sometimes will makes the user feel frustrated and deprive the user of the willing of using the systems again. As for some commercial systems, most of them are too expensive to be affordable for ordinary users. Based on these discussions, this thesis presents a low-cost eye-tracking system. Via the system, the coordinates of the user’s pupil can be computed and then the user’s eye gaze information can be estimated. The eye gaze information is then used to manipulate the movements of the mouse. The system first converts a color image obtained form a CCD camera to a binary image. Using a special projection method we can roughly locate the position of the pupil. The Sobel operator is then used to compute the edges of the pupil. A simple clustering algorithm is adopted to find a circle to fit the edges so that the coordinates of the center of the pupil are located. Finally, a special mapping method is used to transfer the coordinates of pupil to the eye gaze information. Then the eye gaze information is used to manipulate the computer. Several experiments were conducted to test the performance of the proposed system. Experimental results demonstrate the system can quickly and precisely track the movements of the pupil. In addition, via a special application software the system can help the disabled to effectively manipulate the computer to make them be able to live more independently.