燕山大学学报2025，Vol.49Issue(5)：377-394,18.DOI:10.3969/j.issn.1007-791X.2025.05.001
基于深度学习的数字图像水印综述
A review of digital image watermarking based on deep learning
摘要
Abstract
Digital image watermarking technology,aimed at protecting copyright,imperceptibly embeds watermark information into digital images.The extensive application of deep learning in image-related tasks has drawn significant research attention towards the study of digital image watermarking.The research background of digital image watermarking is introduced,with the development history of traditional digital image watermarking research organized along the embedding domain of the watermark.A systematic summary of deep image watermarking research from 2018 to 2023 is provided,focusing on methods to enhance the imperceptibility of digital watermarks and strategies to counter various attacks.Existing deep image watermarking methods are evaluated,and the future research hotspots in the field of deep image watermarking are analyzed.关键词
版权保护/数字水印/深度学习/不可感知性/鲁棒性Key words
copyright protection/digital watermarking/deep learning/imperceptibility/robustness分类
信息技术与安全科学引用本文复制引用
刘宇涵,李冲冲,李博阳,王佳炎,赵静,唐勇..基于深度学习的数字图像水印综述[J].燕山大学学报,2025,49(5):377-394,18.基金项目
河北省省级科技计划项目(246Z0105G) （246Z0105G）
河北省创新能力提升计划项目(22567626H) （22567626H）
