Abstract: Iris location is a very important pretreatment process in the iris recognition system. The speed and accuracy of iris location are required to achieve the performance of iris recognition system. According to the characteristics of the human eye, combining with the geometric method, this paper proposes a fast and accurate iris localization algorithm. It detects a point inside the pupil using the gray-scale features within the pupil before the boundary detection mask is employed to locate three non-collinear boundary points along the inner and outer boundaries respectively. These points constitute two right triangles. According to the Thales’ theorem, the hypotenuse of the right triangle is the diameter of its circumcircle, and then the iris parameters can be calculated. A large number of experiments show that the algorithm in terms of speed and accuracy is improved compared to the classic location algorithm.

Key words: iris location, boundary detection mask, geometry

摘要： 虹膜定位是虹膜识别系统中重要的预处理过程，其定位精度和速度决定了虹膜识别系统的性能。就此，根据人眼的特征，结合几何方法，提出了一种快速准确的虹膜定位算法。该算法根据瞳孔内部灰度特点，在瞳孔内找到一点。利用边缘检测模版，依据该点，分别在虹膜内边界和外边界上找到各自不共线的三个边界点，构成两个直角三角形。根据几何原理，这两个直角三角形的斜边分别为内外边界圆的直径，从而定位虹膜边界参数。大量实验证明，该算法在速度和精度上相比经典定位算法均有提高。

关键词: 虹膜定位, 边缘检测模版, 几何方法