摘要
在介绍现有虹膜识别算法的基础上,针对其局限性,提出一种改进的虹膜定位算法,实现了精确而快速的定位。在该算法中,通过阈值化和Freeman链码进行内边界定位;Canny算子和圆形Hough变换被用于外边界定位;上下眼皮的去除利用Canny算子和线性Hough变换实现,睫毛和镜面反射通过阈值化的方法加以去除。实验验证了该算法的合理性和有效性。
For the purpose of localizing iris precisely and quickly, an improved iris localization algorithm is proposed while based on introducing some prevailing localization algorithms. In this method, inner boundary are localized by thresholding the images and Freeman＇s chain code. Canny operator and circle Hough transform are used to localize outer boundary, and the eyelid is removed by Canny operator and linear Hough transform. The method of thresholding is used to remove the eyelash and mirror reflection. Experimental result shows that our algorithm is rational and feasible.
出处
《电视技术》
北大核心
2006年第7期77-79,共3页
Video Engineering
基金
河北省教育厅自然科学项目(2004124)
