Displacement Detection and Variation Tracking of Bangudae Petroglyphs in Daegok-ri, Ulju, based on Deep Learning

doi:10.69756/jnh.2024.9.3.213

All Issue

2024 Vol.9, Issue 3 Preview Page Next Page

Article

Displacement Detection and Variation Tracking of Bangudae Petroglyphs in Daegok-ri, Ulju, based on Deep Learning 딥러닝 기반 울주 대곡리 반구대 암각화 변위 탐지 및 변화량 추적: 이상윤^1†ㆍ신동준²
Lee, Sang-Yun^1†ㆍShin, Dong-Jun²; ¹한국전자통신연구원 디지털융합연구소 과학치안공공ICT연구센터 책임연구원
²충남대학교 전파정보통신공학과 학사과정 4학년

¹Principal Researcher, Police Science & Public Safety ICT Research Center, Digital Convergence Research Laboratory, Electronics and Telecommunications Research Institute
²BS, Department of Radio Information and Communication Engineering, Chungnam National University

31 December 2024. pp. 213-227

PDF

Abstract

This study utilizes a deep learning architecture to detect displacement variations in the Petroglyphs of Bangudae Terrace in Daegok-ri, Ulju, designated as a National Treasure of South Korea. Despite being a priority candidate for UNESCO World Heritage listing, the Petroglyphs of Bangudae Terrace have faced continuous deterioration due to various environmental factors. To preserve them effectively, this research applies the PiDiNet deep learning model to detect and analyze displacement variations in specific joints of the Petroglyphs of Bangudae Terrace. Analyzing six years of accumulated data revealed a total of 96 cases exceeding the margin of error. Additional experiments using synthetic displacement images achieved an anomaly detection accuracy of 93.3%. This study demonstrates that deep learning-based technical management can play a crucial role in preserving cultural heritage and provides meaningful foundational data for exploring technical approaches to heritage protection in the future.

Keywords

Petroglyphs of Bangudae terrace

Deep learning

Displacement variation detection

PiDiNet

Cultural heritage

본 연구는 대한민국 국보로 지정된 울주 대곡리 반구대 암각화의 변위 변화량 탐지를 위해 딥러닝 아키텍처를 활용한다. 반구대 암각화는 유네스코 세계문화유산 우선 등재 후보로 선정되었지만, 다양한 환경적 요인으로 인해 지속적으로 훼손되고 있다. 이를 효과적으로 보존하기 위해, 본 연구는 PiDiNet 딥러닝 모델을 활용하여 반구대 암각화 특정 절리의 변위 변화량을 탐지 및 분석한다. 6년간 축적된 데이터를 기반으로 변위 변화량의 추이를 분석한 결과, 총 96건의 오차범위 초과 사례를 확인하였다. 가상변위 이미지를 활용한 추가 실험에서는 93.3%의 이상 상황 감지 정확도를 달성하였다. 이 연구는 딥러닝 기반의 기술적 관리가 문화유산 보존에 중요한 역할을 할 수 있음을 실증적으로 제시하며, 향후 문화유산 보호를 위한 기술적 접근 방안을 모색하는 데 유의미한 기초 자료로 활용될 수 있을 것이다.

키워드

울주 대곡리 반구대 암각화

딥러닝

변위 변화량 탐지

PiDiNet

문화유산

References

Jeon, H.T. (2021). "The Significance and Value of Bangudae Petroglyphs in Ulsan," Journal of History and Culture, pp. 3-30.
10.18347/hufshis.2021.80.3
Korea Heritage Service (2019). Daegokcheon Stream Petroglyph Comprehensive Research Report, Daejeon, Korea.
Hwang, S.I., Park, K.G., Yoon, S.O. (2010). "Periodic Immersion of the Bangudae Petroglyphs and Rock Weathering Characteristics," Journal of the Korean Geographical Society, pp. 342-359.
Shin, E.J. (2016)."Preservation and assignment at Petroglyphs of Bangudae," The Art History Journal, pp. 127-141.
10.24828/AHJ.47.127.141
Stefanidi, E., Partarakis, N., Zabulis, X., Adami, L., Ntoa, S., Papagiannakis, G., et al. (2022). "Transferring Traditional Crafts from the Physical to the Virtual World: An Authoring and Visualization Method and Platform," JOCCH, pp.1-24.
10.1145/3484397
Pan, J., Li, L., Yamaguchi, H., Hasegawa, K., Thufail, F.I., Tanaka, S., et al. (2022). "3D reconstruction of borobudur reliefs from 2D monocular photographs based on soft-edge enhanced Deep Learning," ISPRS, pp. 439-450.
10.1016/j.isprsjprs.2021.11.007
Siountri, K., Anagnostopoulos, C.N. (2023). "The Classification of Cultural Heritage Buildings in Athens Using Deep Learning Techniques," Journal of Remote Sensing, pp. 3673-3705.
10.3390/heritage6040195
Bahrami, M., Albadvi, A. (2024). "Deep Learning for Identifying Iran's Cultural Heritage Buildings in Need of Conservation Using Image Classification and Grad-CAM," JOCCH, pp. 1-20.
10.1145/3631130
Kevseroglu, O., Kurban, R. (2024). "Re-exploring the Kayseri Culture Route by Using Deep Learning for Cultural Heritage Image Classification," AICCONF, pp. 196-201.
10.1145/3660853.3660913
Choi, W.W., Park, J.Y., Lee, S.Y. (2021). "Detection of Abnormalities in Wooden Cultural Properties Using the F-anogan Model," IEIE, pp. 168-169.
Lee, S.Y., Park, J.Y. (2022). "Ensemble Deep Learning Model for Prediction of Tilt Displacement forImmovable Cultural Heritage." IEIE, pp. 1397-1398.
An, J.S., Lee, S.Y. (2023). "Prediction of Roof Displacement for Anguk-dong byeolgung using EfficientNet," IEIE, pp. 1046-1049.
Kang, S.J., Lee, S.Y. (2023). "A study on Performance Improvement of Heunginjimun Roof Displacement Determination Model Using Artificial Intelligence Algorithm," IEIE, pp. 1050-1054.
Lee, S.Y., Lim, M.K. (2024). "A Study on the Usefulness of Training Data for Displacement Detection of the Bangudae Petroglyphs in Daegok-ri, Ulsan, based on PiDiNet," IEIE, pp. 1680-1682.
Lee, S.Y., Shin, D.J. (2024). "A Study on Displacement Detection of the Bangudae Petroglyphs in Ulsan Daegok-ri Based on Deep Learning," IEIE, pp. 710-712.
Park, J.Y., Lee, S.Y. (2022). "Displacement Detection of Wooden Cultural Properties Using Unsupervised Learning," ICCE-Asia, pp. 302-305.
10.1109/ICCE-Asia57006.2022.9954670
Lee, S.Y., Kang, S.J. (2023). "Experiments and Analyses of Inclination Prediction for the Heunginjimun Roof based on Deep Learning according to Environmental Integration," SCHDP, pp. 167-179.
10.69756/jnh.2023.8.3.167
Lee, S.Y., An, J.S. (2023)"Experiment and Analysis of Tilt Prediction for Anguk-dong Byeolgung Roof based on EfficientNet according to Integration by Environment," SCHDP, pp. 197-210.
10.69756/jnh.2023.8.3.197
Poma, X., Sappa, A., Humanante, P., Akbarinia, A. (2023). "Dense Extreme Inception Network for Edge Detection," Pattern Recognition, pp. 109461.
10.1016/j.patcog.2023.109461
Su, Z., Liu, W., Yu, Z., Hu, D., Liao, Q., Tian, Q., et al. (2021). "Pixel Difference Networks for Efficient Edge Detection," ICCV, pp.5117-5127.
10.1109/ICCV48922.2021.00507
Liu, Y., Yao, J., Lu, X., Xie, R., Li, L. (2019). "Deep Crack: A Deep Hierarchical Feature Learning Architecture for Crack Segmentation", Neurocomputing, pp. 139-153.
10.1016/j.neucom.2019.01.036
Ye, Y., Yi, R., Gao, Z., Zhu, C., Cai, Z., Xu, K. (2023). "NEF: Neural Edge Fields for 3D Parametric Curve Reconstruction from Multi-view Images," CVPR, pp. 8486-8495.
10.1109/CVPR52729.2023.00820
Zhou, C., Huang, Y., Pu, M., Guan, Q., Huang, L., Ling, H. (2023). "The Treasure Beneath Multiple Annotations: An Uncertainty-aware Edge Detector," CVPR, pp. 15507-15517.
10.1109/CVPR52729.2023.01488
Lin, Q., Li, W., Zheng, X., Fan, H., Li, Z. (2023). "Deep CrackAT: An Effective Crack Segmentation Framework based on Learning Multi-scale Crack Features," IFAC, pp. 106876.
10.1016/j.engappai.2023.106876
Jeon, H.T. (2012). "Early Art in the Prehistoric Age and Ancient Times in Korea and the Bangudae Petroglyphs," History & the Boundaries, pp. 1-47.
Song, H.S. (2016). "Establishment of Digital Content of Korea Petroglyph and Restoration of Life Style of the Prehistoric Age," Humanities Contents, pp. 41-71.
10.18658/humancon.2016.12.43.41
Russell, B.C., Torralba, A., Murphy, K.P., Freeman, W.Y. (2008). "LabelMe: A Database and Web-Based Tool for Image Annotation," IJCV, pp. 157-173.
10.1007/s11263-007-0090-8

Information

Publisher :National Heritage Disaster Prevention Society
Publisher(Ko) :국가유산방재학회
Journal Title :Journal of the Society of Cultural Heritage Disaster Prevention
Journal Title(Ko) :저널국가유산
Volume : 9
No :3
Pages :213-227
DOI :https://doi.org/10.69756/jnh.2024.9.3.213

[1] Jeon, H.T. (2021). "The Significance and Value of Bangudae Petroglyphs in Ulsan," Journal of History and Culture, pp. 3-30.
10.18347/hufshis.2021.80.3

[2] Korea Heritage Service (2019). Daegokcheon Stream Petroglyph Comprehensive Research Report, Daejeon, Korea.

[3] Hwang, S.I., Park, K.G., Yoon, S.O. (2010). "Periodic Immersion of the Bangudae Petroglyphs and Rock Weathering Characteristics," Journal of the Korean Geographical Society, pp. 342-359.

[4] Shin, E.J. (2016)."Preservation and assignment at Petroglyphs of Bangudae," The Art History Journal, pp. 127-141.
10.24828/AHJ.47.127.141

[5] Stefanidi, E., Partarakis, N., Zabulis, X., Adami, L., Ntoa, S., Papagiannakis, G., et al. (2022). "Transferring Traditional Crafts from the Physical to the Virtual World: An Authoring and Visualization Method and Platform," JOCCH, pp.1-24.
10.1145/3484397

[6] Pan, J., Li, L., Yamaguchi, H., Hasegawa, K., Thufail, F.I., Tanaka, S., et al. (2022). "3D reconstruction of borobudur reliefs from 2D monocular photographs based on soft-edge enhanced Deep Learning," ISPRS, pp. 439-450.
10.1016/j.isprsjprs.2021.11.007

[7] Siountri, K., Anagnostopoulos, C.N. (2023). "The Classification of Cultural Heritage Buildings in Athens Using Deep Learning Techniques," Journal of Remote Sensing, pp. 3673-3705.
10.3390/heritage6040195

[8] Bahrami, M., Albadvi, A. (2024). "Deep Learning for Identifying Iran's Cultural Heritage Buildings in Need of Conservation Using Image Classification and Grad-CAM," JOCCH, pp. 1-20.
10.1145/3631130

[9] Kevseroglu, O., Kurban, R. (2024). "Re-exploring the Kayseri Culture Route by Using Deep Learning for Cultural Heritage Image Classification," AICCONF, pp. 196-201.
10.1145/3660853.3660913

[10] Choi, W.W., Park, J.Y., Lee, S.Y. (2021). "Detection of Abnormalities in Wooden Cultural Properties Using the F-anogan Model," IEIE, pp. 168-169.

[11] Lee, S.Y., Park, J.Y. (2022). "Ensemble Deep Learning Model for Prediction of Tilt Displacement forImmovable Cultural Heritage." IEIE, pp. 1397-1398.

[12] An, J.S., Lee, S.Y. (2023). "Prediction of Roof Displacement for Anguk-dong byeolgung using EfficientNet," IEIE, pp. 1046-1049.

[13] Kang, S.J., Lee, S.Y. (2023). "A study on Performance Improvement of Heunginjimun Roof Displacement Determination Model Using Artificial Intelligence Algorithm," IEIE, pp. 1050-1054.

[14] Lee, S.Y., Lim, M.K. (2024). "A Study on the Usefulness of Training Data for Displacement Detection of the Bangudae Petroglyphs in Daegok-ri, Ulsan, based on PiDiNet," IEIE, pp. 1680-1682.

[15] Lee, S.Y., Shin, D.J. (2024). "A Study on Displacement Detection of the Bangudae Petroglyphs in Ulsan Daegok-ri Based on Deep Learning," IEIE, pp. 710-712.

[16] Park, J.Y., Lee, S.Y. (2022). "Displacement Detection of Wooden Cultural Properties Using Unsupervised Learning," ICCE-Asia, pp. 302-305.
10.1109/ICCE-Asia57006.2022.9954670

[17] Lee, S.Y., Kang, S.J. (2023). "Experiments and Analyses of Inclination Prediction for the Heunginjimun Roof based on Deep Learning according to Environmental Integration," SCHDP, pp. 167-179.
10.69756/jnh.2023.8.3.167

[18] Lee, S.Y., An, J.S. (2023)"Experiment and Analysis of Tilt Prediction for Anguk-dong Byeolgung Roof based on EfficientNet according to Integration by Environment," SCHDP, pp. 197-210.
10.69756/jnh.2023.8.3.197

[19] Poma, X., Sappa, A., Humanante, P., Akbarinia, A. (2023). "Dense Extreme Inception Network for Edge Detection," Pattern Recognition, pp. 109461.
10.1016/j.patcog.2023.109461

[20] Su, Z., Liu, W., Yu, Z., Hu, D., Liao, Q., Tian, Q., et al. (2021). "Pixel Difference Networks for Efficient Edge Detection," ICCV, pp.5117-5127.
10.1109/ICCV48922.2021.00507

[21] Liu, Y., Yao, J., Lu, X., Xie, R., Li, L. (2019). "Deep Crack: A Deep Hierarchical Feature Learning Architecture for Crack Segmentation", Neurocomputing, pp. 139-153.
10.1016/j.neucom.2019.01.036

[22] Ye, Y., Yi, R., Gao, Z., Zhu, C., Cai, Z., Xu, K. (2023). "NEF: Neural Edge Fields for 3D Parametric Curve Reconstruction from Multi-view Images," CVPR, pp. 8486-8495.
10.1109/CVPR52729.2023.00820

[23] Zhou, C., Huang, Y., Pu, M., Guan, Q., Huang, L., Ling, H. (2023). "The Treasure Beneath Multiple Annotations: An Uncertainty-aware Edge Detector," CVPR, pp. 15507-15517.
10.1109/CVPR52729.2023.01488

[24] Lin, Q., Li, W., Zheng, X., Fan, H., Li, Z. (2023). "Deep CrackAT: An Effective Crack Segmentation Framework based on Learning Multi-scale Crack Features," IFAC, pp. 106876.
10.1016/j.engappai.2023.106876

[25] Jeon, H.T. (2012). "Early Art in the Prehistoric Age and Ancient Times in Korea and the Bangudae Petroglyphs," History & the Boundaries, pp. 1-47.

[26] Song, H.S. (2016). "Establishment of Digital Content of Korea Petroglyph and Restoration of Life Style of the Prehistoric Age," Humanities Contents, pp. 41-71.
10.18658/humancon.2016.12.43.41

[27] Russell, B.C., Torralba, A., Murphy, K.P., Freeman, W.Y. (2008). "LabelMe: A Database and Web-Based Tool for Image Annotation," IJCV, pp. 157-173.
10.1007/s11263-007-0090-8

Journal National Heritage ISSN:3022-9200(Print) 3022-9383(Online) 저널 국가유산

All Issue