Mural Image Restoration Based on Wavelet-Domain Feature Decomposition and Attention Mechanisms.

Saved in:

Bibliographic Details
Title:	Mural Image Restoration Based on Wavelet-Domain Feature Decomposition and Attention Mechanisms.
Authors:	Zhang, Sen¹ zs1002141365@outlook.com, Maimaitiaili, Gulijiamali² gulijiamali@xjnu.edu.cn, Dawuken, Ayiman³ ayimandawken@gmail.com
Source:	IAENG International Journal of Applied Mathematics. Jul2026, Vol. 56 Issue 7, p2873-2887. 15p.
Subjects:	Mural art, Discrete wavelet transforms, Contours (Cartography), Image reconstruction
Abstract:	Mural image restoration is challenging due to disrupted structures, complex textures, and strong stylistic constraints. To address these issues, this paper proposes a structure-guided restoration framework that integrates line-map guidance, wavelet-domain structure-texture decomposition, and mural-oriented style-aware feature modulation. Line maps provide explicit contour priors for damaged regions, while wavelet decomposition separates low-frequency structural information from high-frequency texture details to improve the coordinated recovery of global layout and local appearance. A mural-oriented style perception module is further introduced to enhance color consistency and stylistic coherence. Experiments on the DhMurals1714 dataset demonstrate that the proposed method achieves competitive performance, with clearer advantages under coarse-line and large-area free-form damage. Under coarse-line masks with a 30% mask ratio, the proposed method achieves 30.13 dB PSNR, 0.9398 SSIM, and 0.0458 LPIPS. These results verify the effectiveness of combining explicit structural guidance with frequency-aware feature modeling for mural image restoration. [ABSTRACT FROM AUTHOR]
	Copyright of IAENG International Journal of Applied Mathematics is the property of International Association of Engineers (IAENG) and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database:	Engineering Source

Description
Abstract:	Mural image restoration is challenging due to disrupted structures, complex textures, and strong stylistic constraints. To address these issues, this paper proposes a structure-guided restoration framework that integrates line-map guidance, wavelet-domain structure-texture decomposition, and mural-oriented style-aware feature modulation. Line maps provide explicit contour priors for damaged regions, while wavelet decomposition separates low-frequency structural information from high-frequency texture details to improve the coordinated recovery of global layout and local appearance. A mural-oriented style perception module is further introduced to enhance color consistency and stylistic coherence. Experiments on the DhMurals1714 dataset demonstrate that the proposed method achieves competitive performance, with clearer advantages under coarse-line and large-area free-form damage. Under coarse-line masks with a 30% mask ratio, the proposed method achieves 30.13 dB PSNR, 0.9398 SSIM, and 0.0458 LPIPS. These results verify the effectiveness of combining explicit structural guidance with frequency-aware feature modeling for mural image restoration. [ABSTRACT FROM AUTHOR]
ISSN:	19929978