Ancient mural super-resolution reconstruction based on conditional diffusion model for enhanced visual information
Source abstract
Abstract
Abstract Ancient murals are cultural treasures with high research value, yet few are well preserved. The intricate textures within murals pose substantial challenges for super-resolution reconstruction. To address issues such as detail loss, color shifts, and inadequate noise control in mural super-resolution reconstruction, this study introduces a mural conditional diffusion model (MCDM) for enhanced image reconstruction. The model integrates three core modules: a residual feature distillation network for feature encoding and detail extraction, a residual self-attention module to enhance global consistency, and a Kolmogorov–Arnold-based implicit representation module for high-frequency detail reconstruction. Furthermore, this study establishes three training schemes across two datasets. Experiments show that MCDM performs best on the mixed dataset, with a PSNR of 23.3 dB and an SSIM of 0.8399. Tests across scenarios show that MCDM ensures smooth images while preserving details. Transfer learning further enhances performance by reducing noise and restoring fine features, providing guidance for future work.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Advanced Image Processing Techniques · Advanced Vision and Imaging · Image and Signal Denoising Methods
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (43)
Show source-provided references
1. Pan, X., Ge, Q. & Pan, J. Damage to ancient mural paintings and petroglyphs caused by Pseudonocardia sp.—a review. Acta Microbiol. Sin. 55, 813–818 (2015).
2. Veneranda, M. et al. In-situ multianalytical approach to analyze and compare the degradation pathways jeopardizing two murals exposed to different environments (Ariadne House, Pompeii, Italy). Spectrochim. Acta Part A 203, 201–209 (2018).10.1016/j.saa.2018.05.115 ↗
3. Ogura, D. et al. Influence of environmental factors on deterioration of mural paintings in Mogao Cave 285, Dunhuang. Case Stud. Build. Rehabil. 13, 105–159 (2021).10.1007/978-3-030-49202-1_6 ↗
4. Xu, Y., Li, J., Song, H. & Du, L. Single-image super-resolution using panchromatic gradient prior and variational model. Math. Probl. Eng. 2021, 9944385 (2021).
5. Huang, Y. et al. Single image super-resolution via multiple mixture prior models. IEEE Trans. Image Process. 27, 5904–5917 (2018).10.1109/TIP.2018.2860685 ↗
6. Kim, J., Lee, J. K. & Lee, K. M. Accurate image super-resolution using very deep convolutional networks. In Proc. IEEE Conference on Computer Vision and Pattern Recognition 1646–1654 (IEEE, 2016).10.1109/CVPR.2016.182 ↗
7. Zhang, D., Shao, J., Li, X. & Shen, H. Remote sensing image super-resolution via mixed high-order attention network. IEEE Trans. Geosci. Remote Sens. 59, 5183–5196 (2020).10.1109/TGRS.2020.3009918 ↗
8. Dai, T. et al. Second-order attention network for single image super-resolution. In Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition, Los Angeles, CA, USA 11065–11074 (IEEE, 2019).10.1109/CVPR.2019.01132 ↗
9. Xiao, C. et al. AM-ESRGAN: super-resolution reconstruction of ancient murals based on attention mechanism and multi-level residual network. Electronics 13, 3142 (2024).10.3390/electronics13163142 ↗
10. Zhang, Y. et al. Image super-resolution using very deep residual channel attention networks. In Proc. European Conference on Computer Vision (ECCV) (eds Ferrari, V., Hebert, M. Sminchisescu, C. & Weiss, Y.) 286–301 (Springer, 2018).10.1007/978-3-030-01234-2_18 ↗
11. Guo, Y. et al. Closed-loop matters: dual regression networks for single image super-resolution. In Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition 5407–5416 (IEEE, 2020).10.1109/CVPR42600.2020.00545 ↗
12. Dong, C., Loy, C. C., He, K. & Tang, X. Image super-resolution using deep convolutional networks. IEEE Trans. Pattern Anal. Mach. Intell. 38, 295–307 (2015).10.1109/TPAMI.2015.2439281 ↗
13. Shi, W. et al. Real-time single image and video super-resolution using an efficient sub-pixel convolutional neural network. In Proc. IEEE Conference on Computer Vision and Pattern Recognition 1874–1883 (IEEE, 2016).10.1109/CVPR.2016.207 ↗
14. Lim, B. et al. Enhanced deep residual networks for single image super-resolution. In Proc. IEEE Conference on Computer Vision and Pattern Recognition Workshops 136–144 (IEEE, 2017).10.1109/CVPRW.2017.151 ↗
15. Namhyuk, A., Byungkon, K. & And Kyung-Ah, S. Fast, accurate, and lightweight super-resolution with cascading residual network. In Proc. European Conference on Computer Vision (ECCV) (eds Ferrari, V., Hebert, M. Sminchisescu, C. & Weiss, Y.) 252–268 (Springer, 2018).
16. Xu, Z., Yan, J. & Zhu, H. Mural image super resolution econstruction based on multi-scale residual attention network. Laser Optoelectron. Prog. 57, 152–159 (2020).
17. Rad, M. S. et al. Srobb: targeted perceptual loss for single image super-resolution. In Proc. IEEE/CVF International Conference on Computer Vision 2710–2719 (IEEE, 2019).10.1109/ICCV.2019.00280 ↗
18. Zhang, W., Liu, Y., Dong, C. & Qiao, Y. Ranksrgan: generative adversarial networks with ranker for image super-resolution. In Proc. IEEE/CVF International Conference on Computer Vision 3096–3105 (IEEE, 2019).10.1109/ICCV.2019.00319 ↗
19. Ledig, C. et al. Photo-realistic single image super-resolution using a generative adversarial network. In Proc. IEEE Conference on Computer Vision and Pattern Recognition 4681–4690 (IEEE, 2017).10.1109/CVPR.2017.19 ↗
20. Lugmayr, A., Danelljan, M., Van Gool, L. & Timofte, R. Srflow: Learning the super-resolution space with normalizing flow. In Computer Vision—ECCV 2020. Lecture Notes in Computer Science 12360 (eds Vedaldi, A., Bischof, H., Brox, T. & Frahm, J. M.) (Springer, Cham, 2020).10.1007/978-3-030-58558-7_42 ↗
21. Cao, J., Jia, Y., Yan, M. & Tian, X. Superresolution reconstruction method for ancient murals based on the stable enhanced generative adversarial network. EURASIP J. Image Video Process. 2021, 1–23 (2021).10.1186/s13640-021-00569-z ↗
22. Ren, H., Sun, K., Zhao, F. & Zhu, X. Dunhuang murals image restoration method based on generative adversarial network. Herit. Sci. 12, 39 (2024).10.1186/s40494-024-01159-8 ↗
23. Kong, Z. et al. Diffwave: a versatile diffusion model for audio synthesis. In Proc.The 9th International Conference on Learning Representations (ICLR) (OpenReview, 2021).
24. Yue, Q. et al. Denoising diffusion probabilistic model for face sketch-to-photo synthesis. IEEE Trans. Circuits Syst. Video Technol. 34, 10424–10436 (2024).10.1109/TCSVT.2024.3409184 ↗
25. Sohl-Dickstein, J., Weiss, E., Maheswaranathan, N. & Ganguli, S. Deep unsupervised learning using nonequilibrium thermodynamics. In Proc. International Conference on Machine Learning (ICML) (eds Bach, F. & Blei, D.) 2256–2265 (PMLR, 2015).
26. Batzolis, G., Stanczuk, J., Schönlieb, C.-B. & Etmann, C. Conditional image generation with score-based diffusion models. arXiv preprint arXiv:2111.13606. (2021).
27. Ho, J., Jain, A. & Abbeel, P. Denoising diffusion probabilistic models. NeurIPS 33, 6840–6851 (2020).
28. Li, H. et al. Srdiff: single image super-resolution with diffusion probabilistic models. Neurocomputing 479, 47–59 (2022).10.1016/j.neucom.2022.01.029 ↗
29. Saharia, C. et al. Image super-resolution via iterative refinement. IEEE Trans. Pattern Anal. Mach. Intell. 45, 4713–4726 (2022).
30. Li, S., Li, S. & Zhang, L. Hyperspectral and panchromatic images fusion based on the dual conditional diffusion models. IEEE Trans. Geosci. Remote Sens. 61, 1–15 (2023).
31. Dong, W. et al. ISPDiff: Interpretable Scale-Propelled Diffusion model for hyperspectral image super-resolution. IEEE Trans. Geosci. Remote Sens. 62, 3407967 (2024).
32. Liu, T. et al. Super-resolution reconstruction of ultrasound image using a modified diffusion model. Phys. Med. Biol. 69, 125026 (2024).10.1088/1361-6560/ad4086 ↗
33. Wu, Z. et al. Super-resolution of brain MRI images based on denoising diffusion probabilistic model. Biomed. Signal Process. Control. 85, 104901 (2023).10.1016/j.bspc.2023.104901 ↗
34. Agustsson, E. & Timofte, R. NTIRE 2017 challenge on single image super-resolution: dataset and study. In Proc. IEEE Conference on Computer Vision and Pattern Recognition Workshops 126–135 (IEEE, 2017).10.1109/CVPRW.2017.150 ↗
35. Vaswani, A. Attention is all you need. NeurIPS 30, 5998–6008 (2017).
36. Cunningham, H. et al. Sparse autoencoders find highly interpretable features in language models. arXiv preprint arXiv:2309.08600 (2023).
37. Chen, Y., Liu, S. & Wang, X. Learning continuous image representation with local implicit image function. In Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition 8628–8638 (IEEE, 2021).10.1109/CVPR46437.2021.00852 ↗
38. Gao, S. et al. Implicit diffusion models for continuous super-resolution. In Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition 10021–10030 (IEEE, 2023).10.1109/CVPR52729.2023.00966 ↗
39. Liu, Z. et al. KAN: Kolmogorov–Arnold networks. arXiv preprint arXiv:2404.19756 (2024).
40. Wu, Y. & He, K. Group normalization. In Proc. European Conference on Computer Vision (ECCV) (eds Ferrari, V. & Hebert, M., Sminchisescu, C. & Weiss, Y.) 3–19 (Springer, 2018).
41. Liu, J., Tang, J. & Wu, G. Residual feature distillation network for lightweight image super-resolution. In Proc. European Conference on Computer Vision (ECCV) (eds Bartoli, A., Fusiello, A.) 41–45 (Springer, 2021).10.1007/978-3-030-67070-2_2 ↗
42. Alvarez-Ramos, V., Ponomaryov, V. & Sadovnychiy, S. Image super-resolution via wavelet feature extraction and sparse representation. Radioengineering 27, 603 (2018).10.13164/re.2018.0602 ↗
43. Xie, C. et al. Large Kernel distillation network for efficient single image super-resolution. In Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition 1283–1292 (IEEE, 2023).10.1109/CVPRW59228.2023.00135 ↗