【作者】 张祥德； 曹宇； 王琪； 刘贤艳； 姚翠莉； 高龙成；

【Author】 Zhang Xiangde Cao Yu Wang Qi Liu Xianyan Yao Cuili Gao Longcheng (School of Sciences Northeastern University Shenyang 110004 China)

【机构】 东北大学理学院；

【摘要】 受人脑工作机制的启发,提出基于差分融合的虹膜识别方法。首先对两幅归一化的虹膜图像分别计算行、列一阶差分;然后,取第一幅图像的行一阶差分、列一阶差分,融合产生一个特征编码矩阵;再次,取第一幅图像的列一阶差分、第二幅图像的行一阶差分,融合产生一个新的特征编码矩阵。计算两个特征编码矩阵的归一化距离,以此判断两幅虹膜图像的异同。在中国科学院自动化所虹膜数据库上实验,经过285290次匹配,算法的正确识别率达到99.91%。更多还原

【Abstract】 Inspired by the working mechanism of human brain,fusion of differences is presented for iris recognition.It calculates the 1st-order differences separately from the rows and columns of two normalized iris image.One characteristic matrix is calculated from rows and columns differences of an iris image.Then,the other characteristic matrix is further calculated from the 1st-order differences of row of one iris image and the column of the other.Then,whether the two iris images are the same or not is judged by normalized distance of the two characteristic matrixes.This method was tested on CASIA Iris Image Database,matched for 285,290 times.The rate of correct recognition is up to 99.91%.更多还原