Deep Learning Applications in Digital Image Security
Saqib Ali Nawaz, Jingbing Li, Uzair Aslam Bhatti, Muhammad Usman Shoukat, Raza Muhammad Ahmad · 2023
Since the rapid change of digital info technology, digital image data has been commonly used in communication and the representation of data. However, this type of information is frequently subject to illegal tampering. Consequently, the problem we are facing is how to guarantee that the copyright of digital images is not demolished. This is a constant issue in the field of digital information. Digital watermarking is a copyright protection scheme that, to some extent, can resolve conflicts regarding the ownership of intellectual property in images. This has significant implications for the study of digital image technology. This article provides a comprehensive review of the recent accomplishments made in this research field by academics from both the United States and other countries. First, a basic module, basic features, and dimension pointers of image watermarking are presented; second, the features of existing digital image watermarking algorithms are evaluated, and the classification and introduction of the application of watermarking knowhow in other fields are provided; third, for the resistance performance of watermarking algorithms, a summary of the robust attacks that existing watermarking algorithms can effectively withstand is provided; fourth, we give a thorough analysis of watermarking strategies used in deep learning settings; and finally, for the resistance behavior of watermarking algorithms, a summary of the robust attacks that existing watermarking.