摘要
如何有效地定位虹膜在图象中的位置,是虹膜识别的关键问题之一。论文提出一种基于主动轮廓线的方法——Snake-Daugman(SD)法——来定位虹膜的边界,先用灰度投影的方法检测出瞳孔内的一点作为瞳孔的伪圆心,该点不要求一定在瞳孔中心附近,只要能落在瞳孔内部即可;然后以该点为中心,在其周围等间隔地取几个点作为初始的snake,按照snake的运行机制不断进化,直到虹膜的内边界为止;接着,计算进化后的snake的形心和snake上的控制点与该形心的距离,取其平均值作为瞳孔的半径,形心作为瞳孔的圆心,即可准确定位出虹膜内边界的位置;最后,以瞳孔的圆心为圆心,以瞳孔的半径为虹膜外边界的初始搜索半径,按照简化了的Daugman方法定位虹膜的外边界。实验表明,与常见的定位方法——Hough变换和Daugman的圆形检测算子——相比,文中的方法速度快,精度高,而且,该方法对瞳孔初始的伪圆心的要求并不高,鲁棒性更强。
It is one of the key problems in iris recognition that how to locate iris's position in an image.This paper
presents a new method of iris location based on active contour--Snake- Daugman (SD).First,a point in pupil called
pseudo-center,which is not required to be located at the real center of pupil,is detected by a method of gray projection; Second, several points spaced evenly around the pseudo-center are chosen as initial snake,and then the snake e-
volves until it gets the interior boundary of iris;Third,two values are computed --one is the real center of pupil,
which is expressed through the center of snake,and the other is the radius of pupil,the average of distances between every control point of snake and the center.Finally,the exterior boundary of iris is located by simplified Daugman's method.Experiments show that the method have better performance both in speed and in precision than Hough transformation and Daugman's method.Moreover,it is more robust because the pseudo-center is not rigorous.
出处
《计算机工程与应用》
CSCD
北大核心
2003年第34期104-107,共4页
Computer Engineering and Applications
基金
辽宁省自然科学基金(编号:2001101022)
辽宁省教育厅高校科研项目(编号:202062040)
