基于LGBP与Bloom滤波器的可撤销掌纹模板生成方法

为了实现对用户生物特征信息的有效保护,提高掌纹身份认证系统的安全性,提出一种掌纹可撤销模板生成方法。首先通过Gabor滤波器获得掌纹数据不同方向、不同尺度的幅值特征,对其提取局部均匀模式LBP特征,然后将二值化的特征直方图序列使用Bloom滤波器进行多对一映射,最后进行不可逆变换,得到可撤销掌纹模板。理论分析和实验结果表明,该方法不仅可以有效保护掌纹特征,而且在密钥丢失时,也具有较高的识别率。

Cancelable palmprint templates based on LGBP and Bloom filters

In order to effectively protect users'' biometric information and improve the security and privacy of palm identity verification system, a method is proposed to generate cancelable palmprint templates. The Gabor magnitude features of palmprint were extracted by multi-resolution and multi-orientation Gabor filters. After obtaining the uniform LBP pattern histogram sequence of Gabor magnitude pictures, Bloom filters were used to achieve many-to-one mapping and position scrambling. Then a cancelable palmprint template was generated after irreversible transformation. Both theoretical analysis and experimental results show that this method can not only protect the original palmprint feature, even the key is stolen or shared, it remains high recognition rate.