İHA tabanlı fotogrametrik veriler ile derin öğrenme destekli yol çatlaklarının otomatik tespiti

Öz

Yol üstyapısının en üst tabakasında meydana gelen yüzey çatlakları, zamanında müdahale edilmediğinde yapısal bozulmalara ve bakım maliyetlerinin artmasına neden olmaktadır. Özellikle köprü yüzeyleri gibi kritik bileşenlerde çatlakların erken tespiti, yapısal bütünlüğün korunması açısından büyük önem taşımaktadır. Geleneksel denetim yöntemleri ise zaman alıcı, maliyetli ve genellikle öznel değerlendirmelere dayalıdır. Bu çalışma, çatlak tespiti sürecinde hem görsel hem de konumsal doğruluk sağlayan yeni nesil bir yaklaşım sunmak üzere, İnsansız Hava Aracı (İHA) görüntülemesi, fotogrametrik modelleme ve derin öğrenme tekniklerini entegre eden bütüncül bir yöntem önermektedir. Kırsal bir yol kesiminde toplanan yüksek çözünürlüklü görüntüler fotogrametri iş akışı ile değerlendirilmiştir. Yapılan doğruluk analizinde iç yöneltme (kamera kalibrasyonu) hatası, dengeleme sonrasında 1.83 mm olarak belirlenmiştir. Dış doğruluk (üç boyutlu (3B) konumsal hassasiyet) ise 2.14 mm Kök Ortalama Kare Hata (Root Mean Square Error – RMSE) değeriyle doğrulanmıştır. Elde edilen ortomozaik görüntüler üzerinde, Transformatör tabanlı CT-CrackSeg kullanılarak otomatik çatlak tespiti gerçekleştirildi. Çatlak maskeleri, saha gözlemleri ve referans ölçümlerle karşılaştırılmış; %92.5 Kesinlik (Precision), %88.3 Duyarlılık (Recall), %90.3 F1-Skoru ve %87.6 Birleşim Üzerinden Kesişim (Intersection over Union – IoU) metrik değerleri elde edilmiştir. Çalışma sonuçları, önerilen yöntemin özellikle düşük trafikli ve yapısal riski yüksek yüzeylerde hızlı, hassas ve tekrar üretilebilir çatlak tespiti için etkin bir çözüm sunduğunu göstermektedir. Yüksek çözünürlüklü, koordinatlı görüntülerin derin öğrenme ile bütünleşik kullanımı, çatlakların hem morfolojik hem de mekânsal analizi açısından literatürde öne çıkan güçlü bir uygulama senaryosu ortaya koymaktadır. Bu yaklaşım, yol bakım yönetim sistemlerine entegre edilerek, proaktif ve veri temelli karar destek mekanizmalarının gelişimine katkı sağlayabilecek niteliktedir.

Anahtar Kelimeler

• Derin öğrenme
• Doğruluk analizi
• İHA Fotogrametrisi
• Segmentasyon
• SfM
• Yol çatlağı tespiti

Deep learning-assisted automatic road crack detection from UAV-based photogrammetric data

Abstract

Surface cracks in pavement structure, if not addressed in a timely manner, can lead to structural deterioration and increased maintenance costs. Early detection of cracks, particularly on critical components such as bridge surfaces, is essential for preserving structural integrity. However, traditional inspection methods are time-consuming, costly, and often rely on subjective assessments. This study proposes an integrated, next-generation approach for crack detection that combines Unmanned Aerial Vehicle (UAV) imaging, photogrammetric modeling, and deep learning techniques to ensure both visual and positional accuracy. High-resolution images collected along a rural road section were evaluated using a photogrammetry workflow. In the accuracy analysis performed, the internal orientation (camera calibration) error was determined to be 1.83 mm after balancing. External accuracy (three-dimensional (3D) positional accuracy) was verified with a Root Mean Square Error (RMSE) value of 2.14 mm. On the resulting orthomosaic images, automatic crack detection was performed using Transformer-based CT-CrackSeg. The predicted crack masks were validated against field observations and reference measurements, yielding 92.5% Precision, 88.3% Recall, 90.3% F1-Score, and 87.6% Intersection over Union (IoU). The results demonstrate that the proposed method provides a reliable, repeatable, and practical solution for fast and accurate crack detection, particularly in low-traffic and structurally sensitive environments. The integration of high-resolution, georeferenced imagery with deep learning enables both morphological and spatial analysis of cracks, offering a powerful use case in current literature. This approach is well-suited for integration into road maintenance management systems and can support the development of proactive, data-driven decision support mechanisms.

Keywords

• Deep learning
• Accuracy analysis
• UAV photogrammetry
• Segmentation
• SfM
• Crack detection

Kaynakça

1. [1] Chen X, Wang C, Liu C, Zhu X, Zhang Y, Luo T, Zhang J. “Autonomous crack detection for mountainous roads using UAV inspection system”. Sensors, 24(14), 4751, 2024.
2. [2] Sadeghi P, Goli A. “Investigating the impact of pavement condition and weather characteristics on road accidents”. International Journal of Crashworthiness, 29(6), 973–989, 2024.
3. [3] Wang W, Zhang Y. “Cylindrical panoramic image stitching based on SIFT algorithm in photogrammetry systems”. Proceedings of the 4th Asia-Pacific Artificial Intelligence and Big Data Forum, 298–303, December 2024.
4. [4] Alkaabi K, El Fawair AR. “Application of a drone camera in detecting road surface cracks: A UAE testing case study”. The Arab World Geographer, 24(3), 221–239, 2021.
5. [5] Kulambayev B, Nurlybek M, Astaubayeva G, Tleuberdiyeva G, Zholdasbayev S, Tolep A. “Real-time road surface damage detection framework based on mask R-CNN model”. International Journal of Advanced Computer Science and Applications, 14(9), 2023.
6. [6] Kırbaş U, Karaşahin M, Ünal EN, Komut M, Demir B, Öcal K. “Development of pavement performance prediction model for bituminous hot mix asphalt on interurban road networks”. Pamukkale Univ Muh Bilim Derg, 23(6), 718–725, 2017. https://doi.org/10.5505/pajes.2016.63497.
7. [7] Munawar HS, Hammad AW, Haddad A, Soares CAP, Waller ST. “Image-based crack detection methods: A review”. Infrastructures, 6(8), 115, 2021.
8. [8] Jia Yi T, Ahmad AB. “Quality assessments of unmanned aerial vehicle (UAV) and terrestrial laser scanning (TLS) methods in road cracks mapping”. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 48, 183–193, 2023.

9. [9] Zhang F, Hu Z, Liang Y, Li Q. “Evaluation of surface crack development and soil damage based on UAV images of coal mining areas”. Land, 12(4), 774, 2023.
10. [10] Azam A, Alshehri AH, Alharthai M, El-Banna MM, Yosri AM, Beshr AA. “Applications of terrestrial laser scanner in detecting pavement surface defects”. Processes, 11(5), 1370, 2023.
11. [11] Siafali E, Tsioras PA. “An innovative approach to surface deformation estimation in forest road and trail networks using unmanned aerial vehicle real-time kinematic-derived data for monitoring and maintenance”. Forests, 15(1), 212, 2024.
12. [12] Zhao Y, Zhou L, Wang X, Wang F, Shi G. “Highway crack detection and classification using UAV remote sensing images based on CrackNet and CrackClassification”. Applied Sciences, 13(12), 7269, 2023.
13. [13] Hong Z, Yang F, Pan H, Zhou R, Zhang Y, Han Y, Liu J. “Highway crack segmentation from unmanned aerial vehicle images using deep learning”. IEEE Geoscience and Remote Sensing Letters, 19, 1–5, 2021.
14. [14] Fakhri SA, Saadatseresht M, Varshosaz M, Zakeri H. “Evaluation of UAV photogrammetric capability in road pavement cracks detection”. Amirkabir Journal of Civil Engineering, 54(5), 1705–1730, 2022.
15. [15] Popa V, Năstase G, Dragomir G, Brezeanu A, Şerban A. “Evaluation of the quality of the road infrastructure with the photogrammetry technique”. IOP Conference Series: Materials Science and Engineering, 1138(1), 012037, April 2021.
16. [16] Kim B, Cho S. “Image-based concrete crack assessment using mask and region-based convolutional neural network”. Structural Control and Health Monitoring, 26(8), e2381, 2019.
17. [17] Cățeanu M, Miclescu SM. “The potential of close-range photogrammetry in evaluating the severity of road surface deformations”. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 48, 77–83, 2024.
18. [18] Ioli F, Pinto A, Pinto L. “UAV photogrammetry for metric evaluation of concrete bridge cracks”. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 43, 1025–1032, 2022.
19. [19] Pan S, Yoshida K, Nishiyama S. “Utilising smartphone-derived photogrammetry 3D model for AI-based top-surfaced asphalt-paved cracks-based instance segmentation and size measurement”. Digital Water, 3(1), 1–20, 2025.
20. [20] Jing J, Ding L, Yang X, Feng X, Guan J, Han H, Wang H. “Topology-informed deep learning for pavement crack detection: Preserving consistent crack structure and connectivity”. Automation in Construction, 174, 106120, 2025.
21. [21] Jing Z, Yanzhi L, Zhongyu J, Siyuan X. “Multi-region segmentation pavement crack detection method based on deep learning”. International Journal of Pavement Research and Technology, 18(1), 56–66, 2025.
22. [22] Agyei Kyem B, Asamoah JK, Aboah A. “Context-CrackNet: A Context-Aware Framework for Precise Segmentation of Tiny Cracks in Pavement Images”. Construction and Building Materials, 484, 141583, 2025.
23. [23] Pascucci N, Alicandro M, Zollini S, Dominici D. “Improving infrastructure monitoring: UAV-based photogrammetry for crack pattern inspection”. Proceedings of the Future Technologies Conference, 351–373, November 2024.
24. [24] Komi D, Yoshida D, Kameyama T. “Development of an automated crack detection system for port quay walls using a small general-purpose drone and orthophotos”. Sensors, 25(14), 4325, 2025.
25. [25] Savino P, Graglia F, Scozza G, Di Pietra V. “Automated corrosion surface quantification in steel transmission towers using UAV photogrammetry and deep convolutional neural networks”. Computer‐Aided Civil and Infrastructure Engineering, 40(14), 2050–2070, 2025.
26. [26] Chen T, Cai Z, Zhao X, et al. “Pavement crack detection and recognition using the architecture of SegNet”. Journal of Industrial Information Integration, 18, 100–144, 2020.
27. [27] Cruz JE, Calsina H, Huacasi L, Mamani W, Beltran N, Puma F, Yana-Mamani V, Huaquipaco S. “Photogrammetry-driven detection of structural peeling in joints and corners of rigid pavements using an unsupervised learning meta-architecture”. IEEE Access, 13, 48132–48145, 2025.
28. [28] Zhao S, Kang F, Li J. “Intelligent segmentation method for blurred cracks and 3D mapping of width nephograms in concrete dams using UAV photogrammetry”. Automation in Construction, 157, 105145, 2024.
29. [29] Liu Z, Lin Y, Cao Y, Hu H, Wei Y, Zhang Z, Lin S, Guo B. “Swin Transformer: Hierarchical vision transformer using shifted windows”. 2021 IEEE/CVF International Conference on Computer Vision (ICCV), 9992–10002, 2021.
30. [30] Liu H, Miao X, Mertz C. “CrackFormer: Transformer network for fine-grained crack detection”. 2021 IEEE/CVF International Conference on Computer Vision (ICCV), 3763–3772, 2021.
31. [31] Xie E, Wang W, Yu Z, Anandkumar A, Álvarez JM, Luo P. “SegFormer: Simple and efficient design for semantic segmentation with transformers”. Neural Information Processing Systems, 2021.
32. [32] Chen X, Shi Y, Pang J. “SECrackSeg: A High-Accuracy Crack Segmentation Network Based on Proposed UNet with SAM2 S-Adapter and Edge-Aware Attention”. Sensors, 25(9), 2642, 2025.
33. [33] Wang W, Zhang Y. “Cylindrical panoramic image stitching based on SIFT algorithm in photogrammetry systems”. Proceedings of the 4th Asia-Pacific Artificial Intelligence and Big Data Forum, 298–303, December 2024.
34. [34] Morita MM, Carvajal DAL, Bagur ILG, Bilmes GM. “A combined approach of SfM-MVS photogrammetry and reflectance transformation imaging to enhance 3D reconstructions”. Journal of Cultural Heritage, 68, 38–46, 2024.
35. [35] Pandey V, Mishra SS. “A review of image-based deep learning methods for crack detection”. Multimedia Tools and Applications, 2025. https://doi.org/10.1007/s11042-025-20729-x.
36. [36] Yüzkat M, İlhan HO, Aydın N. Derin öğrenme temelli nesne tespit yöntemleri kullanılarak insan sperm hücrelerinin tespiti. Pamukkale Üniversitesi Mühendislik Bilimleri Dergisi. 2024;30:482–493.
37. [37] Yurtsever U, Evirgen H, Avunduk MC. “Detection of colon cancer using K-means and deep learning algorithms on histopathological images”. Pamukkale Univ Muh Bilim Derg, 31(5), 0–0, 2025. https://doi.org/10.5505/pajes.2025.71508.
38. [38] Khalid NSK, Savaş S. “A deep learning approach for assessing consumer sentiment in Amazon dataset”. Pamukkale Univ Muh Bilim Derg, 31(7), 0–0, 2025. https://doi.org/10.5505/pajes.2025.45753.
39. [39] Dede MA, Genç Y. “Direct pose estimation from RGB images using 3D objects”. Pamukkale Univ Muh Bilim Derg, 28(2), 277–285, 2022. https://doi.org/10.5505/pajes.2021.08566.
40. [40] Tao H, Liu B, Cui J, Zhang H. “A convolutional-transformer network for crack segmentation with boundary awareness”. 2023 IEEE International Conference on Image Processing (ICIP), 86–90, 2023.
41. [41] Qi W, Ma F, Zhao G, Liu M, Ma J. “CrackSegMamba: A lightweight Mamba model for crack segmentation”. 2024 IEEE International Conference on Robotics and Biomimetics (ROBIO), 601–607, December 2024.