基于V型检测算子和边缘梯度的虹膜卷缩轮检测方法

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.

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

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 extracted the end point coordinates of the internal texture of the frill in the normalized image. Then, determined the approximate position area of the iris frill by curve fitting. Finally, used edge gradient operator to extracted 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.