A Novel Robust Digital Watermarking Technique Using Compressive Sensing for Biometric Data Protection

Biometric systems are those which utilize or are capability of utiliz ing characteristics of human for enrolment, verification or authentication. Biometrics based authentication system that use physiological and/or behavioral traits likes fingerprint, face, signature, palm print and voice are accept and use by many organizat ion now days. But these biometric authentication systems are vulnerable against different attacks when biometric data transfer from one check point to other check point. In this paper, by combined compressive sensing theory and the correlation approach of White Gaussian Noise based robust digital watermarking technique is proposed for biometric embedding. The technique is divided into five steps: generation of linear measurement from watermark biometric data, embedding linear measurement into another biomet ric data, extraction of secure biometric from linear measurement, compressive sensing recovery process of biometric data and recognition of reconstructed biometric data with enrol biometric data for user authenticity. The proposed technique is providing mo re security and more payload capacity compare to traditional watermarking technique in spatial domain. The proposed technique is evaluated based on parameter like Peak Signal to Noise Ratio and Similarity Factor between reconstruct biometric data and enrol led biometric data. This proposed technique is increased overall data storage capacity in term secure enrolled biometric data of biometric authentication system.