首页> 外文会议>Pattern Recognition, 2009. CCPR 2009 >Protecting Biometric Templates Using LBP-Based Authentication Watermarking
【24h】

Protecting Biometric Templates Using LBP-Based Authentication Watermarking

机译:使用基于LBP的身份验证水印保护生物特征模板

获取原文

摘要

In this paper, a novel scheme to protect biometric templates using LBP-based authentication watermarking is proposed. Firstly, the m-LSB (a generalization of LSB) of biometric templates is set to zeros. Then, the modified biometric templates are partitioned into non-overlapping n X n image blocks. The LBP feature of these image blocks can be obtained, and the authentication watermark bits are generated from these features. A series of message patterns are generated according to authentication watermark bits. In order to resist vector quantization (VQ) attack and collage attack, each message pattern is embedded into the m-LSB of another image block which is selected by a position sequence. Theoretic analysis and experimental results have proved that the proposed algorithm maintains recognition rate in the template-based authentication system. And it not only has superior localization and security, but also tolerate low-noise pollution while keep sensitive to malicious manipulations.
机译:本文提出了一种基于LBP的认证水印保护生物模板的新方案。首先,将生物特征模板的m-LSB(LSB的概括)设置为零。然后,将修改后的生物特征模板划分为不重叠的n X n图像块。可以获取这些图像块的LBP特征,并从这些特征中生成认证水印位。根据认证水印位生成一系列消息模式。为了抵抗矢量量化(VQ)攻击和拼贴攻击,每个消息模式都嵌入到另一个由位置序列选择的图像块的m-LSB中。理论分析和实验结果证明,该算法在基于模板的认证系统中保持了识别率。它不仅具有优越的本地化和安全性,而且在对恶意操作保持敏感的同时还可以承受低噪声污染。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号