Biometric data security using Fractional Fourier Transform and chaotic theory

In the recent past, biometrics found an extensive use in the field of data security and access control to maintain data confidentiality and restrict unauthorised access. As data theft cases are increasing the need of securing biometric data is a major concern. This paper presents an efficient and lossless encryption scheme based on transformation and chaotic domain to achieve high level of data confidentiality and security. Unlike conventional methods, the proposed scheme uses the substitution and permutation in reverse order using different domains to provide adequate security level. Experimental results shows that proposed method reduces the peak signal to noise ratio significantly making the proposed algorithm resistant to perceptual attacks. Also an increased key space is achieved due to the use of transformation domain in conjunction with spatial domain. Experimental results also show that proposed method is highly resistant to statistical and crypt analytical attacks which make it suitable for real time applications.