🤖 AI Summary
Existing watermarking methods primarily target low-resolution images and suffer from limited embedding capacity, making it challenging to simultaneously achieve high capacity, high imperceptibility, and strong robustness for high-resolution images. To address this limitation, this work proposes a block-wise embedding strategy that integrates a reversible symmetric encoder-decoder network—comprising three convolutional and deconvolutional layers—with a global–local weighted joint loss function. The approach enables high-capacity watermark embedding of 32,768 bits (0.0313 bpp) in 1024×1024 images. By overcoming the constraints of fixed-length embedding schemes, the method significantly enhances visual quality and noise robustness, demonstrating its effectiveness and superiority in high-capacity watermarking scenarios.
📝 Abstract
Most existing watermarking techniques are primarily designed for low-resolution images, with few methods tailored for high-resolution images. Moreover, the embedding capacity is often limited to fixed lengths (e.g., 30, 100, 256 bits, etc.), which struggles to meet practical demands. To address these issues, this paper proposes a high-capacity robust watermarking method for high-resolution images, capable of embedding a watermark of 4 KB (32,768 bits) into images with a resolution of 1024*1024, achieving an embedding rate of 0.0313 bpp. Specifically, this paper adopts a block-wise strategy to effectively embed the watermark into local regions, enabling the network to train and learn normally even under low-resource conditions. The encoder and decoder structures respectively employ a reversible symmetric architecture with three convolutional and three deconvolutional layers, ensuring consistency in the coupling and decoupling of the watermark and image features. Additionally, the loss function combines global and local losses with weighted contributions. By incorporating constraints on the visual quality and robustness of local block regions, the overall imperceptibility and robustness of the image are further enhanced. Extensive experimental results verify that the proposed method is effective and feasible in high-resolution image scenarios with high-capacity watermarking, while demonstrating strong robustness against various noise attacks.