Multiwavelet-based digital watermarking with support vector machine technique

Due to the multiwavelet offers better smoothness, symmetry, orthogonality than wavelet, using the coherence among neighborhood coefficients in a discrete multiwavelet transform (DMWT), a multiwavelet domain watermarking scheme based on support vector regression (SVR) is proposed. Then, using the good learning ability of the SVM, the watermark is embedded or extracted by adjusting the value between the selected coefficients and the actual output of the trained by SVR. Experimental results show that the proposed algorithm has good image perceptual quality, which also possesses high watermark robustness to common image processing operation and security.