Watermarking algorithm combined with wavelet transform: Research on digital media image information

In this paper, in order to protect digital media image information, a wavelet transform-combined watermarking algorithm was developed. An embedding matrix was obtained by scale-invariant feature transform (SIFT), and the embedding intensity was determined by the fruit fly optimization algorithm (FOA). The embedding and extraction processes of the watermark was analyzed. The designed algorithm was tested using pepper, jet plane, lake, and living room images as examples. The results showed that under attacks such as cropping and rotation, the algorithm had a peak signal-to-noise ratio (PSNR) value more than 60 dB, and its normalized correlation coefficient was more than 0.90. Under different intensities of JPEG compression and salt-and-pepper noise attacks, its stability was good, suggesting a good anti-attack ability. Compared with the fixed embedding strength, using FOA to determine the embedding strength could effectively enhance the performance of the watermarking algorithm. The results verify the effectiveness of the proposed algorithm. It can be applied to protect actual digital media image information.