Visual Image Digital Watermarking Embedding Algorithm Combining 3D Boolean CNN and Arnold Technology

As digital technology has been leaping forward over the past few years, its application in watermarking has become widespread. To enhance the effectiveness of this technology, a novel digital watermarking embedding algorithm based on a 3D Boolean cellular neural network and cat face transformation is proposed in this study. The proposed algorithm is capable of preprocessing the image using cat face transformation and then encrypting it using a three-dimensional (3D) Boolean cellular neural network for the confidentiality of the digital watermark confidentiality. As indicated by the performance comparison experiment, the information entropy and the correlation coefficient of adjacent pixels are better at 7.9993 and 0.0009. The result suggests that Algorithm 4 exhibits strong robustness. Subsequently, in the empirical analysis, the result reveals that the maximum signal-to-noise ratios of R, G, and B in the HSI color space watermark images reach 76.6621, 80.6228, and 77.4841, all at a high level. The above results suggest that this algorithm exhibits good encryption performance. Accordingly, applying this algorithm to the field of visual images can facilitate the development of color image-level authentication and copyright identification.