基于SURF特征的指纹汗孔检测方法
首发时间:2017-12-20
摘要:高分辨率指纹图像汗孔作为第三层特征,具有先天可区分性,因此指纹图像汗孔特征对提高自动指纹识别认证系统安全性有重要作用,而如何准确提取指纹汗孔成为关键一步。传统指纹图像汗孔提取算法基于骨架细化或匹配滤波方法,受算法性能影响,适用性往往不强。本文提出了基于高分辨率指纹图像SURF特征的SVM分类算法。通过提取高分辨率指纹图像SURF特征,进行SVM模型训练与测试,得到指纹图像汗孔位置。本方法基于高分辨率指纹图像数据集。试验表明,该方法对高分辨率指纹图像的汗孔提取准确、鲁棒性好。
关键词: 高分辨率指纹图像 指纹汗孔提取 SURF特征 SVM
For information in English, please click here
Fingerprint pore detection method based on SURF features
Abstract:High-resolution fingerprint pores as level 3 features are inherently distinguishable. The features of fingerprint image pores play an important role in improving the security of automatic fingerprint identification and authentication system, so how to accurately extract fingerprint pores becomes a key step. Traditional algorithm of fingerprint pore extraction based on skeleton refinement or matched filtering. Affected by the algorithm performance, it\'s applicability is not desirable. A SVM classification algorithm based on SURF features of high resolution fingerprint images is proposed in this paper. By extracting SURF features of fingerprint image for SVM training, with trained model to predict and get the final coordinates of fingerprint image pores. Fingerprint pore detection method based on SURF features is tested in high-resolution fingerprint images. Experimental results show that the method of extraction image pores is accurate and robust in fingerprint image pore detection.
Keywords: High-resolution fingerprint image Fingerprint pore extraction SURF feature SVM
基金:
引用
No.****
同行评议
勘误表
基于SURF特征的指纹汗孔检测方法
评论
全部评论0/1000