Abstract: A digital watermarking algorithm that can resist the strong shear attack is proposed. The embedded strategies include: the watermark is embedded into the basis matrix on behalf of the basic characteristics of the host carrier for the first time, and the theoretical analytical result shows that the algorithm may obtain intelligence and resist shear; the watermark is encrypted spread spectrum modulation to enhance the hidden features; the watermark embedding intensity is adaptively adjusted by a comparison in the PSNR (Power Signal-to-Noise Ratio) between the original image and the synthetic image; the improved PCA(Principal Component Analysis)method is introduced, which strengthens the link between the matrix elements, to improve the algorithm’s convergence speed. Experimental results show that the algorithm is robust to compression, filtering, noise as well as strong shear attacks, and confirm the validity of the theoretical result. The algorithm’s embedding capacity is large and host vector may be an image, audio or others. The algorithm is of general applicability and is a blind watermark scheme.

Key words: non-negative matrix factorization, digital watermarking, shear attack, principal component analysis