基于Rubber-sheet模型的虹膜识别算法研究

Research and realization of iris image preprocessing algorithm

为了实现对虹膜内外边界的精确定位,提出了改进的虹膜图像定位算法.虹膜图像预处理过程主要包括虹膜边界定位、归一化处理和图像增强三部分,其中虹膜图像边界定位是实现虹膜识别的重要前提.首先使用质心探测法确定其二值化虹膜图像的内边界中心,然后再对内边界进行曲线拟合,从而实现对虹膜内边界的定位.使用Canny算子进行边缘检测,确定虹膜图像外边界,完成对虹膜边界的定位.使用Rubber-sheet模型对经过定位的虹膜区域图像进行归一化处理,方便虹膜图像的信息提取和编码匹配.实验证明改进的算法有效提高了虹膜图像预处理效率.

In order to achieve the accurate irts boundary locallzation,the image positioning algorithm is put forward.The process of iris image preprocessing mainly includes three parts,iris boundary localization,normalization and image enhancement.The boundary localization is the important prerequisite to achieve iris recognition.Firstly,uses the centroid detection to determine the center of the binary iris image,then does curve fitting for the inner boundary,and get the iris image′s inner boundary.Through Canny operator to determine the iris image′s outer boundary,and completed the localization of iris boundaries.Uses Rubber-sheet model to do normalization process for the localized area of iris image which is convenient of the iris image information extraction and encoding match.Experiments show that the improved algorithm effectively improved the efficiency of iris image preprocessing.