IMPLEMENTATION OF IMAGE AUTHENTICATION USING DIGITAL WATERMARKING WITH BIOMETRIC

Abstract

The rapid global development of E-commerce in terms of digitalization and distribution of digital contents in the form of image, audio, video, increases the possibility of unrestricted duplication and broadcasting of copyrighted data and the protection of crucial documents is highly significant. Digital watermarking inserts watermark into the cover or host data by unnoticeable modification. In this work digital watermarking with biometric features is done. In this work a technique to implement the hiding of an image inside another image using biometric features namely signature and fingerprint using watermarking techniques is done. To accomplish this, a hybrid watermarking scheme consisting of Discrete Wavelet Transform, Discrete Cosine Transform and Singular Value Decomposition (DWT-DCT-SVD) is proposed for image authentication that is robust against attacks. Here, singular values of watermark1 (fingerprint) and watermark2 (signature) are obtained by applying DWT-DCT-SVD. By adding both the singular values of watermarks we acquire the transformed watermark. To improve the security, robustness and provide authenticity for the image, a two-step watermarking method is demonstrated. The evaluation parameters like PSNR (Peak Signal to Noise Ratio), SSIM (Structured Similarity Index Method), normalized correlation coefficient (NCC) are used for image quality assessment.