一种虹膜特征提取与匹配方法

An Iris Feature Extraction and Matching Method

人眼虹膜尺寸很小,并且容易受到干扰,如果不能有效地提取出稳定的虹膜特征进行模式匹配,将严重影响虹膜识别的准确性和鲁棒性。本文提出一种稳定的虹膜特征提取与匹配方法,在虹膜注册端通过构建强特征分类器获得增强型特征模板,在虹膜识别端通过多样本特征映射融合提取稳定的特征样本,并根据风险预测自适应地确定分类阈值,然后进行特征匹配与虹膜分类,从而有效提高虹膜识别的准确性和鲁棒性。

The real size of human iris is very small,and iris imaging is prone to interference. The accuracy and robustness of iris recognition will be strongly impacted if iris stable features are not extracted and matched effectively. An iris stable feature extraction and matching method is proposed,which produces enhanced feature templates in iris enrollment stage by building strong feature classifiers,and structures stable feature samples in iris recognition stage by multi-sample feature mapping and fusion,and adaptively determines the classification threshold by risk prediction,and then implements feature matching and iris classification. The experimental results show that the method can improve the accuracy and robustness of iris recognition efficiently.