摘要
针对虹膜定位存在速度慢、外边缘模糊导致提取困难等问题,提出一种利用灰度直方图增强对比度、粗定位与精定位结合的三步定位算法。先利用灰度投影粗定位出瞳孔圆心,再根据灰度直方图将图像二值化后,用Roberts算子提取边缘,通过粗略搜索缩小搜索范围,实现精确定位.试验结果表明,该算法减少了传统方法中搜索的盲目性,解决了外边缘定位效果不佳的问题,对图像质量要求不高,且快速有效、精度高。
In order to improve the accuracy of locating iris exterior edge and the speed of iris localization,we proposed an improved algorithm based on histogram equalization and coarse-to-fine localization.First,we detect a coarse center of a pupil through the projection of image intensity in a binary image.Then,a fine center of the pupil is found out by using a Roberts mask.Finally,the location of iris' outer edge can be finished in the similar way. Through our experiments,the algorithm of this paper is proved to be valid and efficient.Furthermore,this algorithm is not sensitive to the quality of iris image.
出处
《电子科技大学学报》
EI
CAS
CSCD
北大核心
2007年第S2期1129-1131,共3页
Journal of University of Electronic Science and Technology of China
关键词
虹膜
虹膜定位
虹膜识别
瞳孔
iris
iris localization
iris recognition
pupil
