Multipurpose Image Watermarking Based on the Wavelet Tree Contrast Level Transformation

Abstract

Multipurpose image watermarking algorithms can achieve the goal of content authentication and copyright protection simultaneously. In this paper, we propose a wavelet tree based blind watermarking algorithm using contrast level transformation. The wavelet coefficients of the host image are grouped into trees. The robust watermark and the semi-fragile watermark are embedded by quantizing one tree and resetting the other. With level change compensation, the watermarked image has high fidelity to the original one since the difference is minimized. Both of the watermarks can be independently extracted without the original image. The simulation results have demonstrated the effectiveness of the proposed algorithm in terms of robustness compared with other wavelet tree based technique.