Abstract:With the growing popularity of the Internet, digital images are used and transferred more frequently. Although this phenomenon facilitates easy access to information, it also creates security concerns and violates intellectual property rights by allowing illegal use, copying, and digital content theft. Using watermarks in digital images is one of the most common ways to maintain security. Watermarking is proving and declaring ownership of an image by adding a digital watermark to the original image. Watermarks can be either text or an image placed overtly or covertly in an image and are expected to be challenging to remove. This paper proposes a combination of convolutional neural networks (CNNs) and wavelet transforms to obtain a watermarking network for embedding and extracting watermarks. The network is independent of the host image resolution, can accept all kinds of watermarks, and has only 11 layers while keeping performance. Performance is measured by two terms; the similarity between the extracted watermark and the original one and the similarity between the host image and the watermarked one.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the technical challenges of embedding and extracting watermarks in digital images to protect the copyright of images and prevent illegal use. Specifically, the paper proposes a method that combines Convolutional Neural Networks (CNNs) and Discrete Wavelet Transform (DWT) to achieve this goal. This method aims to improve the robustness of the watermark, that is, the watermark can still be accurately extracted after suffering various attacks, while maintaining the quality of the watermarked image, making it as similar as possible to the original image. ### Main contributions of the paper: 1. **Enhanced robustness**: By introducing an attack simulator, the training network can better resist common attacks such as Gaussian noise, salt - and - pepper noise, and JPEG compression. 2. **Performance optimization**: The proposed network structure has only 11 layers and does not depend on the resolution of the host image, and can accept various types of watermarks. 3. **Quality improvement**: Using DWT as a pre - processing technique improves the quality of the watermarked image, making the watermark more difficult to detect. ### Technical details: - **Discrete Wavelet Transform (DWT)**: It is used to decompose the image into different sub - bands (LL, LH, HL, HH) and select the appropriate sub - band to embed the watermark. - **Convolutional Neural Networks (CNNs)**: They are used for the pre - processing, embedding and extraction processes of images and watermarks. - **Loss function**: It consists of two parts. One is the Mean Squared Error (MSE) which measures the similarity between the watermarked image and the original image, and the other is the Mean Absolute Error (MAE) which measures the similarity between the extracted watermark and the original watermark. The general form of the loss function is: \[ L_3=\lambda_1L_1 + \lambda_2L_2 \] where \(\lambda_1\) and \(\lambda_2\) are hyper - parameters used to adjust the balance between invisibility and robustness. ### Experimental results: - **PSNR**: Peak Signal - to - Noise Ratio, which is used to evaluate the invisibility of the watermarked image. The experimental results show that even under attack, the PSNR value is still high, indicating that the quality of the watermarked image is good. - **BER**: Bit Error Rate, which is used to evaluate the robustness of the watermark. The experimental results show that in the absence of attack, the BER value is very low, indicating that the accuracy of watermark extraction is high. ### Conclusion: The method proposed in the paper has made significant progress in improving watermark robustness and image quality, especially in combating common attacks. Future work can further optimize the attack simulator to make it more general, and at the same time explore other neural network architectures to further improve performance.

Convolutional Neural Network-Based Image Watermarking using Discrete Wavelet Transform