A weight factor based secure invisible image watermarking using Discrete Cosine Transform

The rapid growth of the internet and technology laid the stones for the beginning of the globalization era. The exchange of multimedia data like images, videos, and audio over large networks is increasing more rapidly. The challenge faced here is to protect ownership rights against thefts and illegal authorities. In such cases, it is imperative to develop security tools to conceal multimedia to safeguard ownership rights. In this article, we proposed an invisible image watermarking methodology to protect the copyrights. The original image is transformed using discrete cosine transform into a frequency domain and 128-bit AES is applied to the copyright image following which embedding is performed using an embedding factor α which is estimated using multiple linear regression. We compare the PSNR of watermarked images with respect to the original image and normalized correlation is used to compare the similarity of the extracted images with respect to the copyright images after different attacks. The results show that the proposed watermarking scheme possesses great robustness against all different attacks.