A Novel Biometric Fingerprint Template Based Method for Stream Cipher Design

In biometric-based security systems, generation of random biometric keys is a significant means to realize secure, reliable and robust data confidentiality. Here, a novel method for stream cipher design is proposed using biometric fingerprint template. The biometric feature vector is extracted from user's fingerprint in discrete wavelet transform domain. The feature vector is shuffled and inter-mixed with chaotic sequence to eliminate similar patterns, if any. The whole method of enciphering and deciphering is under the control of user's secret key and fingerprint template. The method is apt enough to yield great encryption effect and can resist cryptographic attacks. The method is tested on digital images to investigate its encryption performance. The measures like histogram distribution, pixels auto-correlation, entropy, plain-image sensitivity are assessed and compared with recent proposals to demonstrate effectiveness, suitability and consistency of proposed method.