A Novel Digital Watermarking Approach against Geometical Distortions using DWT and LS-SVM

This paper proposes a novel algorithm for blind watermarking by applying discrete wavelet transformation and least squares support vector machine into watermark embedding and detection. It is strong against general attacks in the traditional digital watermarking, but difficult to geometric attacks. Firstly, discrete wavelet transformation is performed on coefficient block and watermark image is embedded. Subsequently, the trained least squares support vector machine is employed to extract the watermark blindly. Experimental results confirm that the proposed scheme is not only robust against common attacks such as noise, filter and crop, but also robust against geometrical distortions.