| Summary: In this study, robust watermarking systems for copyright protection of digital images have been mentioned. For this purpose, new watermarking methods using different decomposition techniques based on wavelet transform have been proposed. In the first approach using Discrete Wavelet Transform and Singular Value Decomposition methods, a new version of the Firefly Algorithm, the Self Adaptive Step Firefly Algorithm, has been used for the first time to balance the conflicting imperceptibility and robustness. Since this algorithm determines the next step of each firefly based on its present situation and historical information, it enables a global exploration of the solution space. In other approach, the iris code, which presents an effective solution for proving the ownership of the embedded watermark is hidden into color images. Here, Redistributed Invariant Discrete Wavelet Transform has been used to solve the robustness problem of the Discrete Wavelet Transform against geometric attacks. With this transform technique, different approaches based on Singular Value Decomposition, Schur Decomposition and QR Decomposition optimized by Firefly Algorithm,Self Adaptive Step Firefly Algorithm are proposed. Among these approaches, the most appropriate method with regard to imperceptibility, robustness and authentication accuracy has been tried to determine. Research findings prove the effectiveness of both traditional and biometric based watermarking systems in terms of perceptual transparency and robustness. In addition, when the authentication performance of the biometric watermarking study is examined, it is observed that the results are quite satisfactory.
Key Words: Digital image watermarking, Biometric watermarking, Discrete wavelet transform, Singular value decomposition, QR decomposition, Schur decomposition. |
