Deep Learning-Based Crack Detection on Cultural Heritage Surfaces
This study employs a deep learning-based object detection model, GoogleNet, to identify cracks in cultural heritage images. Subsequently, a semantic segmentation model, SegNet, is utilized to determine the location and extent of the cracks. To establish a scale ratio between image pixels and real-wo...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-07-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/15/14/7898 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849406618442137600 |
|---|---|
| author | Wei-Che Huang Yi-Shan Luo Wen-Cheng Liu Hong-Ming Liu |
| author_facet | Wei-Che Huang Yi-Shan Luo Wen-Cheng Liu Hong-Ming Liu |
| author_sort | Wei-Che Huang |
| collection | DOAJ |
| description | This study employs a deep learning-based object detection model, GoogleNet, to identify cracks in cultural heritage images. Subsequently, a semantic segmentation model, SegNet, is utilized to determine the location and extent of the cracks. To establish a scale ratio between image pixels and real-world dimensions, a parallel laser-based measurement approach is applied, enabling precise crack length calculations. The results indicate that the percentage error between crack lengths estimated using deep learning and those measured with a caliper is approximately 3%, demonstrating the feasibility and reliability of the proposed method. Additionally, the study examines the impact of iteration count, image quantity, and image category on the performance of GoogleNet and SegNet. While increasing the number of iterations significantly improves the models’ learning performance in the early stages, excessive iterations lead to overfitting. The optimal performance for GoogleNet was achieved at 75 iterations, whereas SegNet reached its best performance after 45,000 iterations. Similarly, while expanding the training dataset enhances model generalization, an excessive number of images may also contribute to overfitting. GoogleNet exhibited optimal performance with a training set of 66 images, while SegNet achieved the best segmentation accuracy when trained with 300 images. Furthermore, the study investigates the effect of different crack image categories by classifying datasets into four groups: general cracks, plain wall cracks, mottled wall cracks, and brick wall cracks. The findings reveal that training GoogleNet and SegNet with general crack images yielded the highest model performance, whereas training with a single crack category substantially reduced generalization capability. |
| format | Article |
| id | doaj-art-4c3e6842b9374c9e84ef7050589cdbf3 |
| institution | Kabale University |
| issn | 2076-3417 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj-art-4c3e6842b9374c9e84ef7050589cdbf32025-08-20T03:36:19ZengMDPI AGApplied Sciences2076-34172025-07-011514789810.3390/app15147898Deep Learning-Based Crack Detection on Cultural Heritage SurfacesWei-Che Huang0Yi-Shan Luo1Wen-Cheng Liu2Hong-Ming Liu3Department of Civil and Disaster Prevention Engineering, National United University, Miaoli 360302, TaiwanDepartment of Civil and Disaster Prevention Engineering, National United University, Miaoli 360302, TaiwanDepartment of Civil and Disaster Prevention Engineering, National United University, Miaoli 360302, TaiwanDepartment of Civil and Disaster Prevention Engineering, National United University, Miaoli 360302, TaiwanThis study employs a deep learning-based object detection model, GoogleNet, to identify cracks in cultural heritage images. Subsequently, a semantic segmentation model, SegNet, is utilized to determine the location and extent of the cracks. To establish a scale ratio between image pixels and real-world dimensions, a parallel laser-based measurement approach is applied, enabling precise crack length calculations. The results indicate that the percentage error between crack lengths estimated using deep learning and those measured with a caliper is approximately 3%, demonstrating the feasibility and reliability of the proposed method. Additionally, the study examines the impact of iteration count, image quantity, and image category on the performance of GoogleNet and SegNet. While increasing the number of iterations significantly improves the models’ learning performance in the early stages, excessive iterations lead to overfitting. The optimal performance for GoogleNet was achieved at 75 iterations, whereas SegNet reached its best performance after 45,000 iterations. Similarly, while expanding the training dataset enhances model generalization, an excessive number of images may also contribute to overfitting. GoogleNet exhibited optimal performance with a training set of 66 images, while SegNet achieved the best segmentation accuracy when trained with 300 images. Furthermore, the study investigates the effect of different crack image categories by classifying datasets into four groups: general cracks, plain wall cracks, mottled wall cracks, and brick wall cracks. The findings reveal that training GoogleNet and SegNet with general crack images yielded the highest model performance, whereas training with a single crack category substantially reduced generalization capability.https://www.mdpi.com/2076-3417/15/14/7898deep learningGoogleNetSegNetsegmentationcrack detectioncultural heritage |
| spellingShingle | Wei-Che Huang Yi-Shan Luo Wen-Cheng Liu Hong-Ming Liu Deep Learning-Based Crack Detection on Cultural Heritage Surfaces Applied Sciences deep learning GoogleNet SegNet segmentation crack detection cultural heritage |
| title | Deep Learning-Based Crack Detection on Cultural Heritage Surfaces |
| title_full | Deep Learning-Based Crack Detection on Cultural Heritage Surfaces |
| title_fullStr | Deep Learning-Based Crack Detection on Cultural Heritage Surfaces |
| title_full_unstemmed | Deep Learning-Based Crack Detection on Cultural Heritage Surfaces |
| title_short | Deep Learning-Based Crack Detection on Cultural Heritage Surfaces |
| title_sort | deep learning based crack detection on cultural heritage surfaces |
| topic | deep learning GoogleNet SegNet segmentation crack detection cultural heritage |
| url | https://www.mdpi.com/2076-3417/15/14/7898 |
| work_keys_str_mv | AT weichehuang deeplearningbasedcrackdetectiononculturalheritagesurfaces AT yishanluo deeplearningbasedcrackdetectiononculturalheritagesurfaces AT wenchengliu deeplearningbasedcrackdetectiononculturalheritagesurfaces AT hongmingliu deeplearningbasedcrackdetectiononculturalheritagesurfaces |