Digital Image Watermarking Techniques Using Machine Learning—A Comprehensive Survey

Digital image watermarking is the most interesting and active field for research as it prevents unwanted access to multimedia data. The trade-off between imperceptibility, robustness, capacity and safety must be maintained for the conception of an efficient and strong digital picture watermarking system. Different studies have been conducted in order to ensure that these needs are hybridized by many domains, including spatial and transformational fields. An analytical analysis is performed on existing digital picture watermarking systems in this research. The digital information that has resulted in the request for a safe ownership of the information may recently be readily changed, reproduced, distributed and stored. The watermark solution for the authentication of content and copyright protection is quite good. This paper discusses basic concepts and features of digital watermarking, important attacks on watermarking systems, general embedding and extraction processes for watermarking marks, and important techniques for the transformation using machine learning are analysed. The objective of this paper is to provide an ephemeral study and background on the definition, and idea and major contributions of watermarking the techniques are classified according to different categories: host signal, sensitivity, robustness, kind of watermark, essential data for extraction, processing domain and applications.