Image watermarking is the process of embedding a secret message, watermark, inside an image such that the visual perception of the watermarked image is unaltered and the watermark is invisible and robust to attacks. This secret watermark is used for copyright protection and ownership authentication. The two traditional approaches for image watermarking are the spatial and spectral domain techniques. In the spatial domain, the watermark is embedded in selected regions chosen based on the texture of the given image [1, 2]. While in the spectral domain, the watermark is embedded in the transform domain using methods such as DCT and DWT, in the mid-frequency range to ensure transparency and robustness of the watermark, simultaneously [3]. The DWT remains one of the most effective and easy to implement techniques in image watermarking. It has also been used in various image processing applications such as image denoising.
The biggest issue in DWT-based image watermarking is how to choose the coefficients to embed the watermark. The most common approaches include modifying the largest DWT coefficients in all decomposition levels or quantizing certain DWT coefficients in different levels and scales. Other approaches mark the host image by setting modulo 2 differences between the largest and the smallest coefficients according to the watermark bit value [4, 5]. The effectiveness of DWT-based image denoising in separating the wavelet coefficients that belong to the noise and the signal motivates us to use it for watermarking. The threshold derived for separating the ’significant’ and the ’insignificant’ coefficients, i.e. signal and noise, will be used to determine the coefficients to be watermarked.
