数字图像多功能水印综述

Multipurpose image watermarking aims to provide two or more image protection functionalities by embedding various types of watermarks into the same image, including capabilities of copyright protection, content authentication, and tampered recovery. But it still faces challenges in terms of performance and application scenarios nowadays. To promote the study of multipurpose image watermarking, this paper reviews digital image watermarking first, covering its system characteristics, classifications, key technologies, achievable functions, and application scenarios. Furthermore, multipurpose image watermarking methods developed in the past decade are compared and analyzed in terms of fusion-watermark embedding, zero-watermark embedding, serial-watermark embedding, and parallel-watermark embedding. Finally, the future research prospects and development trends are discussed. It is important to note that embedding multiple watermarks results in performance losses in most cases, and further improvements are needed to enhance its effectiveness in real-world application scenarios. Leveraging artificial intelligence techniques in image multipurpose watermarking holds the promising potential to enhance performance.