苑玮琦,朱立军,张波.基于形态学与支持向量机的虹膜坑洞纹理检测[J].仪器仪表学报,2017,38(3):664-671
基于形态学与支持向量机的虹膜坑洞纹理检测
Iris crypt texture detection based on morphology and SVM
  
DOI:
中文关键词:  坑洞纹理;形态学;区域生长  支持向量机;虹膜识别
英文关键词:crypt texture  morphology  region growing  support vector machine(SVM)  iris identification
基金项目:国家自然科学基金(61271365)项目资助
作者单位
苑玮琦 沈阳工业大学 视觉检测技术研究所沈阳110870 
朱立军 1.沈阳工业大学 视觉检测技术研究所沈阳110870; 2.沈阳化工大学计算机科学与技术学院沈阳110142 
张波 沈阳工业大学 视觉检测技术研究所沈阳110870 
AuthorInstitution
Yuan Weiqi Computer Vision Group,Shenyang University of Technology,Shenyang 110870, China 
Zhu Lijun 1. Computer Vision Group,Shenyang University of Technology,Shenyang 110870, China; 2.School of Computer Science and Technology,Shenyang University of Chemical Technology,Shenyang 110142, China 
Zhang Bo Computer Vision Group,Shenyang University of Technology,Shenyang 110870, China 
摘要点击次数: 665
全文下载次数: 0
中文摘要:
      坑洞纹理是虹膜表面上一种重要的特征纹理。对可见光虹膜图像而言,如何能不受睫毛、眼睑、光斑及光照不均匀等干扰因素的影响,并能快速、准确地提取出该特征纹理仍然是目前一个亟待解决的难题。提出一种基于形态学和支持向量机(SVM)的可见光虹膜坑洞纹理检测方法。首先使用灰度形态学和二值形态学相结合的方法提取出所有目标纹理;然后使用区域生长方法定位所有目标纹理并计算各个目标纹理的特征向量;最后再使用SVM和定义约束条件的方法提取出最终的坑洞纹理。通过实验证明:该方法能较好地克服光斑等干扰的影响,对坑洞状纹理的检出率高于其他同类方法。
英文摘要:
      Crypt texture is an important feature texture on iris surface. For visible light iris image, how to extract the crypt texture fast and accurately without being influenced by the interference factor, such as eyelashes, eyelids, light spot and uneven illumination is still a challenging task. In order to solve this problem, a visible light iris crypt texture detection method based on morphology and support vector machine (SVM) is proposed in this paper. Firstly, the method combining gray scale morphology and two value morphology is applied to extract all the target textures in the iris image; and then, the region growing method is used to locate all the target textures and calculate their corresponding feature vectors; at last, SVM and the method defining constraint condition are applied to extract the final crypt textures. Experiment results show that the proposed method can overcome the interference of light spot and etc., and is superior to other methods in terms of crypt texture detection rate.
查看全文  查看/发表评论  下载PDF阅读器