Super Sesolution Reconstruction of Ancient Paintings

The preservation of ancient paintings is crucial for safeguarding cultural heritage. However, these artifacts often suffer from degradation, necessitating effective digital restoration techniques. In this study, we propose the Multi-scale Feature Fusion Network (MSFFNet) for super-resolution reconstruction of ancient paintings. MSFFNet leverages Multi-scale Feature Fusion (MSFF) blocks to extract and aggregate features at varying scales, enabling the reconstruction of fine details and complex textures. Comprehensive experiments on benchmark datasets demonstrate that MSFFNet outperforms state-of-the-art methods in terms of PSNR and SSIM, achieving superior visual quality. An ablation study further highlights the critical role of MSFF blocks in enhancing reconstruction performance. By bridging the gap between advanced deep learning techniques and digital restoration, our work contributes to the preservation and appreciation of cultural artifacts.