With the ubiquity of mobile computing and the increasing use of cameras in smart devices, multimedia information such as images, audio, and video have become of immense value in our daily lives. Multimedia applications place higher requirements on the security of image information in communication and storage. Therefore, in this paper, a novel zero-watermarking algorithm based on the main features of digital images is proposed. First, the key feature points from the host image are converted into a binary sequence based on a threshold value, and the edge features of a given image are analyzed. Then, every edge feature is taken in the construction process, which enhances the robustness and distinguishability of the scheme. In particular, the resulting image pixel values are used as an initial configuration for the two-dimensional Cellular Automaton. The Cellular Automaton is set up to run for defined generations, thus acquiring the feature image. The scheme uses the behavior characteristics of Cellular Automaton generations which will add diffusion property to the scrambling process. The watermark bits are scrambled using Cellular Automaton generations to enhance the security of the proposed algorithm. Finally, the zero-watermark image is generated by performing XOR manipulation on the feature image and the scrambled watermark information, which will be stored in the intellectual property database as the basis for copyright authentication. The experimental results show that the proposed algorithm is robust against various distortion and geometric rotation attacks as the values of BER and NC are close to or equal to the optimal values of BER and NC, "0 and 1", respectively. Compared with other similar zero watermarking schemes, the proposed scheme has low computational complexity and shows better performance against different types of attacks, such as filtering, adding noise, and geometry.
