A Novel Digital Image Watermark Algorithm Based on Wavelet Fractal Compression

摘要

In the paper, a novel blind watermark algorithm in wavelet domain based on fractal compression to perform watermark embedding and extraction is presented. The fractal encoding is performed on the coefficients at various scales and orientations for a certain wavelet block having self-similarity, and the watermarks sequence is embedded in host image by applying fractal encoding so that the decoded one produces the watermarked image. In addition, the logistic mapping is carried out on the row and column of watermark. What is most important is that the original image isn't necessary in the process of extracting watermark. We tested the algorithm performance in terms of different Stirmark attacking methods, such as JPEG compression, middle-value filtering, noise interference. The experimental results show that the algorithm greatly improved in terms of the watermark robustness and security.