Blind Image Watermarking Scheme Based on Wavelet Tree Quantization Robust to Geometric Attacks

This paper proposes an improved blind image watermarking scheme based on wavelet tree quantization. The wavelet coefficients of the original image are grouped into wavelet super trees. Each watermark bit is embedded using two wavelet super trees. The watermark is extracted according to the statistical difference between two wavelet super trees. In order to better provide a trade-off between the strength of the watermark and quality of the watermarked image, the knowledge of the relation between Discrete Wavelet Transform (DWT) coefficients and Human Visual System (HVS) is applied to the watermarking algorithm. Simulation results show that the proposed algorithm is robust to both common image processing operations and geometric attacks such as rotation, scaling, cropping.