An efficient two pass lossless invisible watermarking algorithm for natural images

In this paper, we propose a novel method for image watermarking which can recover the original image after extracting the embedded data without effecting the original cover image. The proposed embedding algorithm runs in two pass. The watermarked image after first pass works as a cover image for second pass. In each pass, embedded data utilizes different prediction algorithm. First pass uses a highly effective method for prediction whether in second pass a simpler and less complex method is used in our work. Moreover, the proposed algorithm has simple decoder complexity. Our algorithm gives better embedding capacity and Peak Signal to Noise Ratio (PSNR) than various One Pass algorithms. We also modified our two pass algorithm which increases PSNR value by sacrificing payload capacity. © 2012 Institute of Telecommunica.