A Lossless Model for Generation of Unique Digital Code for Identification of Biometric Images

Abstract

Security of biometric data is a common problem in this modern era during transmission and storage of biometric data. This paper proposes zero-watermarking scheme that resolves the problem of biometric data security without tempering the biometric data. The biometric pattern of iris is employed to create digital binary code for iris images. This created binary pattern is combined with watermark which is identity of associated person to create a unique ID. This unique ID is completely uninformative which does not reveal the biometric information as well as person’s identity. At the receiver end, person ID can be recovered by the certmed personnel only on permission of the created unique ID. Experiments are done on the openly accessible CASIA dataset available on iris images and achieved results which are motivating in the path of biometric image identification and certmcation. The proposed work is a novel method which can be used in biometric security applications and can be very helpful to detect identity fraud.