Abstract:Reversible visible watermarking (RVW) is an active copyright protection mechanism. It not only transparently superimposes copyright patterns on specific positions of digital images or video frames to declare the copyright ownership information, but also completely erases the visible watermark image and thus enables restoring the original host image without any distortion. However, existing RVW algorithms mostly construct the reversible mapping mechanism for a specific visible watermarking scheme, which is not versatile. Hence, we propose a generic RVW framework to accommodate various visible watermarking schemes. In particular, we obtain a reconstruction data packet -- the compressed difference image between the watermarked image and the original host image, which is embedded into the watermarked image via any conventional reversible data hiding method to facilitate the blind recovery of the host image. The key is to achieve compact compression of the difference image for efficient embedding of the reconstruction data packet. To this end, we propose regularized Graph Fourier Transform (GFT) coding, where the difference image is smoothed via the graph Laplacian regularizer for more efficient compression and then encoded by multi-resolution GFTs in an approximately optimal manner. Experimental results show that the proposed framework has much better versatility than state-of-the-art methods. Due to the small amount of auxiliary information to be embedded, the visual quality of the watermarked image is also higher.

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is the balance between the universality and visual quality of existing reversible visible watermarking algorithms. Specifically: 1. **Universality Limitations**: Most of the existing reversible visible watermark (RVW) algorithms construct reversible mapping mechanisms (RMR) for specific visible watermarking schemes, which results in limited universality of these methods. That is, these algorithms cannot adapt well to different visible watermarking schemes. 2. **Balance between Visual Quality and Reversibility**: Among the existing RVW methods, it is difficult to achieve both good visual effects and complete reversibility simultaneously. The process of constructing RMR often restricts the flexibility of visible watermarks, thus affecting the visual quality of watermarked images. To address these challenges, the author proposes a general - purpose reversible visible watermarking framework with the following characteristics: - **Universality**: By introducing a reconstruction data packet to assist in removing the visible watermarked image, this framework can be applied to any visible watermarking algorithm. This reconstruction data packet is obtained by compressing the difference image between the watermarked image and the original host image and is embedded into the watermarked image as auxiliary information. - **Visual Quality**: Through efficient compression methods, especially the regularized graph Fourier transform (GFT) - based coding (RGFT), the amount of auxiliary information is small, thus ensuring the visual quality of the watermarked image. - **Blind Extraction**: This method can blindly extract the watermark and restore the original host image without relying on the original visible watermarked image. ### Main Contributions 1. **General - purpose Reversible Visible Watermarking Framework**: A general - purpose RVW framework is proposed. By introducing a reconstruction data packet, a general - purpose reversible mapping relationship (RMR) is constructed, and reversibility can be achieved for any visible watermarking algorithm. 2. **Regularized Graph Fourier Transform Coding (RGFT)**: A coding method based on the regularized graph Fourier transform is proposed for efficiently compressing the difference image, thereby generating a compact reconstruction data packet. This method improves the visual quality of the watermarked image while maintaining reversibility. 3. **Experimental Results**: The experimental results show that the proposed framework is superior to existing methods in terms of universality and compression efficiency. Due to the small amount of embedded auxiliary information, the visual quality of the watermarked image is also higher. ### Technical Details 1. **Data Embedding Process**: - First, embed the visible watermark image \(W\) into the region of interest (ROI) of the host image \(I\) to generate the watermarked image \(I_w\). - Calculate the difference image \(D=(I - I_w)_{\text{ROI}}\). - Compress the difference image \(D\) to generate the reconstruction data packet \(D_c\). - Use traditional reversible data hiding techniques to invisibly embed \(D_c\) into the non - ROI area of \(I_w\) to generate the final watermarked image \(I_w'\). 2. **Image Restoration Process**: - Extract \(D_c\) from the watermarked image \(I_w'\), decompress it and reconstruct the difference image \(D\). - Restore the original host image \(I\) through the inverse process \(I = I_w+D\). 3. **Regularized Graph Fourier Transform Coding (RGFT)**: - Use the graph Laplacian regularizer (GLR) to smooth the difference image, making it more compact in the frequency domain. - Compress the smoothed difference image through the multi - resolution GFT coding method to optimize the rate - distortion performance. ### Formulas - **Graph Laplacian Regularizer**: \[ x^\top L x=\sum_{i \sim j} a_{i,j}\cdot(x_i - x_j)^2 \] where \(i \sim j\) indicates that vertex \(i\) and \(j\) are connected, and \(a_{i,j}\) is the edge weight. - **Graph Fourier Transform**: \[ \hat{x}=U^\top x \] where \(U\) is an orthogonal matrix containing eigenvectors.

Generic Reversible Visible Watermarking Via Regularized Graph Fourier Transform Coding

Object watermarks for digital images and video

Research on Reversible Visible Watermarking Algorithms Based on Vectorization Compression Method

An algorithm for removable visible watermarking

Robust Reversible Watermarking Scheme in Video Compression Domain Based on Multi-Layer Embedding

Multipurpose Video Watermarking Algorithm for Copyright Protection and Tamper Detection

The Algorithm of Content Adaptive Color Image Watermark Based on RDWT and NVF

A Semi-Fragile Blind Watermarking Scheme For Color Images Based On Visual Cryptography And Discrete Cosine Transform

Improved Reversible Visible Watermarking Based On Adaptive Block Partition

Robust Reversible Watermarking Via Clustering and Enhanced Pixel-Wise Masking.

Reversible Visible Watermarking and Lossless Recovery of Original Images

A robust reversible watermarking scheme overcomes the misalignment problem of generalized histogram shifting

A Reversible Watermarking Scheme for High-Fidelity Applications

Reliable Embedding for Robust Reversible Watermarking.

Study on reversible visible watermarking scheme for digital images

A Highly Robust Reversible Watermarking Scheme Using Embedding Optimization and Rounded Error Compensation

A geometrically robust multi-bit video watermarking algorithm based on 2-D DFT

Improved Reversible Visible Image Watermarking Based on HVS and ROI-selection

A Recursive Embedding Algorithm Towards Lossless 2D Vector Map Watermarking

Reversible Visible Watermarhing with Lossless Data Embedding Based on Difference Value Shift.

Research on scalable video watermarking algorithm based on H.264 compressed domain