基于V型检测算子和边缘梯度的虹膜卷缩轮检测方法  

Iris frill detection method based on V-shaped detection operator and edge gradient

作　　者：张波 吕翠翠 Zhang Bo;Lyu Cuicui(School of Computer Science&Technology,Shenyang University of Chemical Technology,Shenyang 110142,China)

机构地区：[1]沈阳化工大学计算机科学与技术学院,沈阳110142

出　　处：《计算机应用研究》2021年第6期1917-1920,共4页Application Research of Computers

基　　金：辽宁省博士科研启动基金资助项目(2019-BS-191)。

摘　　要：针对虹膜卷缩轮位置检测中卷缩轮边界定位困难的问题进行了研究,提出了一种基于V型检测算子和边缘梯度的虹膜卷缩轮检测方法。首先根据虹膜卷缩轮内部纹理的位置特征、形状特征和灰度特征,定义V型检测算子,在归一化图像中,利用V型检测算子对图像进行处理,提取卷缩轮内部纹理终点坐标;其次,通过曲线拟合得到卷缩轮边界的初定位区域;最后利用边缘梯度算子在初定位区域进行检测提取卷缩轮边界。实验结果表明,该方法的正确检出率达到93.2%,可以有效地提取虹膜卷缩轮边界。To solve the problem that difficulty in extracting iris frill,this paper proposed a detection method based on V-shaped detection operator and edge gradient.Firstly,it defined the V-type detection operator based on the position,shape and grayscale characteristics of the internal texture of the frill to extract the end point coordinates of the internal texture of the frill in the normalized image.Then,it determined the approximate position area of the iris frill by curve fitting.Finally,it used edge gradient operator to extract iris frill in the approximate position area.Experimental results show that the positioning accuracy of this method reaches 93.2%,which can effectively extract iris frill.

关 键 词：虹膜卷缩轮 V型检测算子 曲线拟合 归一化图像 边缘梯度算子 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]