Detection of Road Damages Using Machine Learning Methods with Data Collected from Various Geographies: A Study on Türkiye

Year 2024, Volume: 3 Issue: 3, 255 - 270, 31.10.2024

Ahmet Cihangir Kavcı , Ömer Faruk Cansız

https://doi.org/10.62520/fujece.1421398

Abstract

Road damage seriously affects the comfort and safety of drivers. The detection of road damage is of great importance not only for transportation safety, but also in terms of cost. The detection of road damage is critical for enabling early intervention and repair. In this study, the road damage detection performance of the YOLO (You Only Look Once) v8 algorithm was evaluated using datasets obtained from different geographies, including Czechia -Türkiye, India-Türkiye, USA-Türkiye, and Japan-Türkiye. The findings revealed both the capabilities of the algorithm in damage detection and the challenges it faced in distinguishing certain types of damage. For the creation of the Türkiye dataset, images of roads in the province of Hatay were recorded. These images were labeled using Microsoft's VoTT application. Comparisons and evaluations were made among the developed models. Among these models, the Japan-Türkiye model yielded the best results with a 0.55 mAP and 0.54 F1 score. The results of the models indicated that the appearance of damage varies according to the geographical location and the quality of road data. It was observed that data consisting of local images and uncertain damage types were important in training.

Keywords

Road damage detection, YOLO algorithm, Machine learning, Object detection, RDD2022

Ethical Statement

There is no need to obtain ethics committee permission for the prepared article.

Supporting Institution

Scientific Research Projects Coordination Office at İskenderun Technical University

Project Number

2022LTP06

Thanks

This investigation was conducted with the assistance of the Scientific Research Projects Coordination Office at İskenderun Technical University, within the framework of project number 2022LTP06.

Farklı Coğrafyalardan Elde Edilen Verilerle Yol Hasarlarının Makine Öğrenmesi Yöntemleri Kullanılarak Tespiti: Türkiye Üzerine Bir İnceleme

Abstract

Karayolu hasarı, özellikle sürücülerin konforunu ve güvenliğini ciddi şekilde etkilemektedir. Yollardaki hasarların tespiti, sadece ulaşım güvenliği açısından değil, aynı zamanda maliyet açısından da büyük önem taşımaktadır. Yol hasarlarının tespiti, erken müdahale ve onarımı sağlamak açısından kritik öneme sahiptir. Bu çalışmada, YOLO (You Only Look Once) v8 algoritmasının yol hasar tespit performansı, Çekya-Türkiye, Hindistan-Türkiye, ABD-Türkiye ve Japonya-Türkiye dahil olmak üzere farklı coğrafyalardan elde edilen veri setleri kullanılarak değerlendirildi. Bulgular, algoritmanın hasar tespit konusundaki yeteneklerini ve belirli hasar türlerini ayırt etmede karşılaştığı zorlukları ortaya koydu. Türkiye veri setinin oluşturulması için Hatay ilindeki yolların görüntüleri kaydedildi. Bu görüntüler, Microsoft'un VoTT uygulaması kullanılarak etiketlendi. Geliştirilen modeller arasında karşılaştırmalar ve değerlendirmeler yapıldı. Bu modeller arasında en iyi sonuçları Japonya-Türkiye modeli, 0.55 mAP ve 0.54 F1 skoru ile verdi. Modellerin sonuçları, hasarın görünümünün coğrafi konuma ve yol verilerinin kalitesine göre değiştiğini gösterdi. Yerel görüntülerden ve belirsiz hasar türlerinden oluşan verilerin eğitimde önemli olduğu gözlemlendi.

Keywords

Yol hasar tespiti, YOLO algoritması, Makine öğrenmesi, Nesne tespiti, RDD2022

Project Number

2022LTP06

References

• Highway Transportation Statistics, “Karayolu Ulasım İstatistikleri (2021).” [Online]. Available:https://www.kgm.gov.tr/SiteCollectionDocuments/KGMdocuments/Yayinlar/YayinPdf/KarayoluUlasimIstatistikleri2021.pdf
• K. G. M. B. İ. Dairesi, “Bölgeler — kgm.gov.tr.” [Online]. Available: https://www.kgm.gov.tr/Sayfalar/KGM/SiteTr/Bolgeler/Bolgeler.aspx.
• K. G. M. B. İ. Dairesi, “Bolge5 — kgm.gov.tr.” [Online]. Available: https://www.kgm.gov.tr/Sayfalar/KGM/SiteTr/Bolgeler/5Bolge/Harita.aspx.
• H. Maeda, Y. Sekimoto, T. Seto, T. Kashiyama, and H. Omata, “Road Damage Detection and Classification Using Deep Neural Networks with Smartphone Images,” Comput.-Aided Civ. Infrast.. Eng., vol. 33, no. 12, pp. 1127–1141, Dec. 2018.
• W. Wang, B. Wu, S. Yang, and Z. Wang, “Road Damage Detection and Classification with Faster R-CNN,” in 2018 IEEE International Conference on Big Data (Big Data), Seattle, WA, USA: IEEE, Dec. 2018, pp. 5220–5223.
• M.-T. Cao, Q.-V. Tran, N.-M. Nguyen, and K.-T. Chang, “Survey on performance of deep learning models for detecting road damages using multiple dashcam image resources,” Adv. Eng. Inform., vol. 46, p. 101182, Oct. 2020.
• D. Arya et al., “Transfer Learning-based Road Damage Detection for Multiple Countries,” Aug. 2020.
• D. Arya et al., “Global Road Damage Detection: State-of-the-art Solutions,” Nov. 2020.
• D. Arya, H. Maeda, S. K. Ghosh, D. Toshniwal, and Y. Sekimoto, “RDD2020: An annotated image dataset for automatic road damage detection using deep learning,” Data Brief, vol. 36, p. 107133, May 2021.
• D. Jeong and J. Kim, “Road Damage Detection using YOLO with Image Tiling about Multi-source Images,” in 2022 IEEE International Conference on Big Data (Big Data), Dec. 2022, pp. 6401–6406.
• S. Wang et al., “An Ensemble Learning Approach with Multi-depth Attention Mechanism for Road Damage Detection,” in 2022 IEEE International Conference on Big Data (Big Data), Dec. 2022, pp. 6439–6444.
• Z. Lu, L. Ding, Z. Wang, L. Dong, and Z. Guo, “Road Condition Detection Based on Deep Learning YOLOv5 Network,” in 2023 IEEE 3rd International Conference on Electronic Technology, Communication and Information (ICETCI), May 2023, pp. 497–501.
• F. Xie and G. Liang, “Using Highly Accurate and Portable Deep Learning Models in Road Damage Detection and Classification,” in 2022 10th International Conference on Information Systems and Computing Technology (ISCTech), Dec. 2022, pp. 172–175.
• M. Sathvik, G. Saranya, and S. Karpagaselvi, “An Intelligent Convolutional Neural Network based Potholes Detection using Yolo-V7,” in 2022 International Conference on Automation, Computing and Renewable Systems (ICACRS), Dec. 2022, pp. 813–819.
• D. Arya, H. Maeda, S. K. Ghosh, Durga Toshniwal, and Yoshihide Sekimoto, “RDD2022: A multi-national image dataset for automatic Road Damage Detection,” Sep. 2022.
• “70mai Dash Cam 4K A800S,” dashcam.70mai.com. [Online]. Available: https://dashcam.70mai.com/a800s/
• “microsoft/VoTT,” GitHub. [Online]. Available: https://github.com/Microsoft/VoTT
• J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, “You Only Look Once: Unified, Real-Time Object Detection,” in 2016 IEEE Conf. Comput. Vis. Pattern Recognit. CVPR, June 2016.
• G. Jocher, A. Chaurasia, and J. Qiu, “YOLO by Ultralytics,” GitHub. [Online]. Available: https://github.com/ultralytics/ultralytics
• J. Terven, D.-M. Córdova-Esparza, and J.-A. Romero-González, “A Comprehensive Review of YOLO Architectures in Computer Vision: From YOLOv1 to YOLOv8 and YOLO-NAS,” Mach. Learn. Knowl. Extr., vol. 5, no. 4, pp. 1680–1716, Dec. 2023.
• “Google Colaboratory,” colab.research.google.com. [Online]. Available: https://colab.research.google.com