IWT based robust and secure color image watermarking using Hessenberg decomposition and SVD

The use of watermarking techniques is becoming more popular among researchers as a potential option for protecting digital image copyright. In this paper, a secure and robust colour image watermarking scheme using integer wavelet transform (IWT) and dual matrix decomposition is proposed. Singular value decomposition (SVD) and Hessenberg decomposition (HD) techniques are utilized for embedding the watermark. Watermark is pre-processed with IWT and SVD before inserting it into cover image. Principle components (PC’s) of cover image are utilized for embedding the watermark. To provide the security, watermark is encrypted using chaotic encryption method before embedding. Robustness and imperceptibility analysis of proposed technique is assessed. Various parameters are utilized for assessing the performance of proposed scheme. Simulation results shows that the proposed technique provides good imperceptibility and robustness even after applying different types of attacks. Obtained PSNR and SSIM is above 54.9 dB and 0.9992 respectively for all cover images without attacks which implies that the proposed technique offers better visual quality of watermarked images. Robustness analysis implies that this technique is also able to survive various attacks and extract the embedded watermark successfully. Comparison is also performed with existing techniques and it is clear from comparison, that proposed scheme performs better in terms of invisibility and robustness against various attacks.