当前位置: X-MOL 学术Peer-to-Peer Netw. Appl. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Cancelable Iris template for secure authentication based on random projection and double random phase encoding
Peer-to-Peer Networking and Applications ( IF 3.3 ) Pub Date : 2021-01-03 , DOI: 10.1007/s12083-020-01046-6
Vani Rajasekar , J. Premalatha , K. Sathya

Biometric recognition approaches have overwhelmed major issues in traditional authentication systems. Privacy invasion and security are the needed features in the effective development of biometric recognition systems. Recently some biometric techniques have been suffered to hacking trails also. Hence there is a despairing need to propose a novel cancelable iris recognition scheme. The prime objective of this research is to create a transformed cancelable biometric template through an encryption function and a one-way transformation function. The random projection matrix is employed here to generate the first feature vector. Then an encrypted cancelable iris code is generated using Double Random Phase Encryption (DRPE) in the Fractional Fourier Transform (FFT) domain. The proposed framework utilizes both the left and right iris image of a person and generates a single cancelable iris template. This feature guarantees the privacy-preserving cancelable iris code generation for secure authentication. The experiment is carried on two state-of-art datasets CASIA Iris V4 and IITD iris to affirm the efficacy of the proposed methodology. The result analysis witnessed a promising accuracy of 99.59%, a recognition rate of 99.88% with a lower Equal Error Rate (ERR) of 0.46%. It has also been proved that the proposed approach is computationally efficient as it recognizes the iris code with less recognition time of 7 ms with a maximum true positive and true negative rate of 100%.



中文翻译:

可取消的虹膜模板,用于基于随机投影和双重随机相位编码的安全身份验证

生物特征识别方法已经淹没了传统身份验证系统中的主要问题。隐私入侵和安全性是生物识别系统有效开发所必需的功能。最近,一些生物识别技术也遭受了黑客攻击。因此,迫切需要提出一种新颖的可消除虹膜识别方案。这项研究的主要目的是通过加密功能和单向转换功能创建一个可转换的可取消生物识别模板。这里采用随机投影矩阵来生成第一特征向量。然后,使用分数阶傅立叶变换(FFT)域中的双随机相位加密(DRPE)生成加密的可取消虹膜代码。所提出的框架利用人的左和右虹膜图像,并生成单个可取消的虹膜模板。此功能保证了用于安全身份验证的保留隐私的可取消虹膜代码的生成。实验在两个最新的数据集CASIA Iris V4和IITD虹膜上进行,以确认所提出方法的有效性。结果分析的准确性为99.59%,识别率为99.88%,而平均错误率(ERR)较低,为0.46%。还已经证明,所提出的方法在计算上是有效的,因为它以7 ms的较少识别时间识别虹膜代码,最大正负率为100%。此功能保证了用于安全身份验证的保留隐私的可取消虹膜代码的生成。实验在两个最新的数据集CASIA Iris V4和IITD虹膜上进行,以确认所提出方法的有效性。结果分析的准确性为99.59%,识别率为99.88%,而平均错误率(ERR)较低,为0.46%。还已经证明,提出的方法在计算上是有效的,因为它以7 ms的较少识别时间识别虹膜代码,最大正负率为100%。此功能保证了用于安全身份验证的保留隐私的可取消虹膜代码的生成。实验在两个最新的数据集CASIA Iris V4和IITD虹膜上进行,以确认所提出方法的有效性。结果分析的准确性为99.59%,识别率为99.88%,而平均错误率(ERR)较低,为0.46%。还已经证明,所提出的方法在计算上是有效的,因为它以7 ms的较少识别时间识别虹膜代码,最大正负率为100%。46%。还已经证明,提出的方法在计算上是有效的,因为它以7 ms的更少识别时间识别虹膜代码,最大正负率为100%。46%。还已经证明,提出的方法在计算上是有效的,因为它以7 ms的更少识别时间识别虹膜代码,最大正负率为100%。

更新日期:2021-01-03
down
wechat
bug