Digital Watermarking Algorithm Based On Fractal Color Image Coding

Digital watermark technology was proposed for solving the copyright problem. In recent years,the image watermarking based on fractal encoding method has been developed.The traditional fractal color image coding method is mainly for the purpose of image compression, which is discussed in RGB space and applies three gray-level linear predictions to encode the image components independently. In this paper, we propose a new fractal color image coding method, which, different from the traditional method, mainly focuses on improving the quality of the watermarked image. In the method, we consider each pixel of the image as a three dimensional vector; a general form of 3′3 matrix is used as a scaling operator; and the luminance offset vector is replaced by the range block mean vector.We next propose a quantization method for fractal coefficients to guarantee the contractivity requirement of the new fractal color image coding. Then an orthogonalization fractal color coding method is achieved to get very high image quality. Experiments show that the proposed method can perform better than the traditional fractal image coding method in image quality(measured by PSNR).Finally, we put forward a specific algorithm for fractal watermarking of color images. Based on theoretical analysis, we know that the orthogonalization fractal color decoding is a mean vector-invariant iteration. the range block mean vector is a good place for hiding watermark. Meanwhile, according to the characteristics of human visual system(HVS), we implement the embedding process in the CIE* *La b space and utilize a just noticeable difference(JND) profile to ensure the invisibility of the watermark. Experimental results show that the proposed method not only can obtain an imperceptible change in image quality, but also has good robustness against various typical attacks.