Construction of a digital restoration model for ancient cultural relics based on HAT-CNN-FCC-SRGAN

To address the common problems of insufficient restoration accuracy and low image resolution in existing digital restoration models for ancient artifacts, a computer vision based digital restoration model for ancient artifacts has been proposed to improve the effectiveness of image restoration of ancient artifacts. The method first uses computer vision technology to accurately recognize and restore the structure and texture information of ancient artifact images. Then, the Generative Adversarial Network enhances the resolution of the restored images, thereby improving the overall restoration quality. The experiment was conducted on a self built dataset of cultural relic images, and the performance of the model was verified by combining objective indicators with subjective evaluation. The results showed that the proposed model achieved peak signal-to-noise ratio of 30.8dB ~ 32.5dB, structural similarity of 0.942–0.956, and perceptual similarity LPIPS of 0.089. Both pixel level and perceptual level indicators were significantly better than mainstream comparison models. The starting distance of Fr é chet is as low as 3.21, which is the best among similar methods. In terms of computational efficiency, the model only takes 72.3 s to repair 100 images and occupies only 135.4 MB of memory, which is significantly better than similar methods. Overall, the model achieves a high-quality balance between restoration accuracy, semantic consistency, and computational efficiency, providing an efficient and feasible technical solution for the digital restoration of ancient artifacts.

Authors

Institutions

Publication Details

Journal
Discover Artificial Intelligence
Published
2026-11-17
DOI
https://doi.org/10.1007/s44163-026-02702-1
Primary Topic
Generative Adversarial Networks and Image Synthesis
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Construction of a digital restoration model for ancient cultural relics based on HAT-CNN-FCC-SRGAN

孙鹏
6 links
Discover Artificial Intelligence
Generative Adversarial Networks and Image Synthesis
article

Construction of a digital restoration model for ancient cultural relics based on HAT-CNN-FCC-SRGAN

孙鹏
article en

Abstract

To address the common problems of insufficient restoration accuracy and low image resolution in existing digital restoration models for ancient artifacts, a computer vision based digital restoration model for ancient artifacts has been proposed to improve the effectiveness of image restoration of ancient artifacts. The method first uses computer vision technology to accurately recognize and restore the structure and texture information of ancient artifact images. Then, the Generative Adversarial Network enhances the resolution of the restored images, thereby improving the overall restoration quality. The experiment was conducted on a self built dataset of cultural relic images, and the performance of the model was verified by combining objective indicators with subjective evaluation. The results showed that the proposed model achieved peak signal-to-noise ratio of 30.8dB ~ 32.5dB, structural similarity of 0.942–0.956, and perceptual similarity LPIPS of 0.089. Both pixel level and perceptual level indicators were significantly better than mainstream comparison models. The starting distance of Fr é chet is as low as 3.21, which is the best among similar methods. In terms of computational efficiency, the model only takes 72.3 s to repair 100 images and occupies only 135.4 MB of memory, which is significantly better than similar methods. Overall, the model achieves a high-quality balance between restoration accuracy, semantic consistency, and computational efficiency, providing an efficient and feasible technical solution for the digital restoration of ancient artifacts.

Discover Artificial IntelligenceVol. 6(1)
Dalian Jiaotong University (CN)
Generative Adversarial Networks and Image Synthesis
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon