摘要
一幅虹膜图象采集后 ,首先要进行质量评价 ,再根据评价结果确定其是否可以用于虹膜识别 .将虹膜图象质量评价分为 3个部分 :虹膜检测、虹膜不可用区域的确定以及虹膜图象清晰度评价 ,即采用模板匹配法在图象中检测可能的虹膜 ,并利用虹膜图象的一些特征来校验其可信度 ;寻找虹膜与眼皮的边界线 ,计算虹膜被眼皮遮盖的面积 ,同时根据虹膜和瞳孔的边界参数来计算位于图象外虹膜面积的大小 ,以确定不可用虹膜部分的大小 ;用灰度差分累加和的方法计算瞳孔边沿的平均高度 ,以判断图象中虹膜的清晰程度 .实验结果表明 ,该方法能够有效地筛选出符合要求的虹膜图象 .
After an Iris image is acquired, its quality should be firstly evaluated to judge whether it can be used for Iris recognition. In this paper Iris image quality evaluation is divided into three parts, namely, Iris detection, Iris area calculating covered by the eyelids or not in the image as well as definition evaluation. The potential Iris is detected with template match and some characteristic Iris features of the Iris image are used for verifying the presence of the Iris. The boundary between the Iris and the eyelids is found to acquire the Iris area covered by eyelids, and then the Iris area not in the image is computed according to the features of Iris boundary and pupil boundary. The average height of pupil boundary is calculated for evaluating the definition of the iris image. The experimental results indicate that the Iris images with high quality can be picked out effectively.
出处
《中国图象图形学报（A辑）》
CSCD
北大核心
2003年第4期387-391,共5页
Journal of Image and Graphics
