近來虹膜辨識已成為身份證明的有效方法之一，一個完整的虹膜辨識系統需要包含五個基本步驟：影像獲取、虹膜分割、正規化、特徵抽取、與配對，其中虹膜分割是決定配對精確與否的關鍵步驟，而瞳孔偵測與定位又是此步驟的首要的工作，許多研究已提出各種方法來偵測瞳孔邊界，由於各自的缺點無法精確地偵測到瞳孔邊界，因此導致在處理辨識上的錯誤，本論文提出結合霍氏轉換(Hough Transform)與積-微分算子(Integro-differential Operator)的一個混合技術以偵測瞳孔邊界，實驗結果顯示本技術在改進瞳孔與虹膜定位的精確性上有好的性能。
Recently, iris recognition is one of effective methods in person identification. A complete iris recognition system must include five basic steps: image acquisition, iris segmentation, normalization, feature extraction and matching. Iris segmentation is the key step to determine the accuracy of matching. Pupil detection and localization is the foremost task in this step. Many researches had proposed several methods to detect the pupil boundary. However, these existing methods are not able to detect pupil boundary accurately due to their own disadvantages and hence leading to errors in identification process. In this thesis a hybrid technique of combining Hough transform and integro-differential operator has developed to detect the boundary of pupil. Experimental results show that the technique has a good performance, which improve the accuracy on pupil and iris localization.
