기계학습 기반의 노면 결함 탐지 시스템
An Automated Road Surface Defect Detection System based on Machine Learning
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
최근 기후 변화로 인해 여름에는 국지적 폭우, 겨울에는 폭설 등의 자연재해가 빈번하게 발생하면서 도로파손이 증가하고 있으며, 이는 주행 차량의 오작동 및 주행 안전성에 심각한 영향을 미쳐 대형 사고로 이어질 위험을 높이고 있다. 현재 도로관리 시스템은 주로 시각적 검사와 노동력을 통해 자료를 수집하고 분석하고 있어 실시간으로 변화하는 도로 상황을 효과적으로 감시하기 어렵다. 따라서 인공지능을 활용하여 도로결함을 실시간으로 검출하고 관찰하는 새로운 기술이 필요하다. 본 논문에서는 노면에서 발생할 수 있는 다양한 결함을 자동으로 검출하는 인공지능 기반 노면 결함 탐지 시스템을 제안한다, 제안된 시스템은 현실 도로에서 발생할 수 있는 18가지 종류의 다양한 노면 결함을 실시간으로 검출하고 인식할 수 있음을 현실 세계의 데이터를 통해 검증하였다.
Abstract
Due to recent climate changes, natural disasters such as localized heavy rain in the summer and heavy snow in the winter have become more frequent, leading to increased road damage. This significantly impacts the malfunction and safety of vehicles, thereby increasing the risk of major accidents. Current road management systems primarily rely on visual inspections and manual labor to collect and analyze data, making it difficult to effectively monitor rapidly changing road conditions in real-time. Therefore, there is a need for new technology that utilizes artificial intelligence to detect and monitor road defects in real-time. This paper proposes an AI-based pavement defect detection system that automatically identifies various defects that can occur on road surfaces. The proposed system has been validated using real-world data to detect and recognize 18 types of road surface defects in real-time.
참고문헌 (11)
- B. Mo, "Competition between shared autonomous vehicles and public transit: A case study in Singapore", Transportation Research Part C, 127, pp.103058~ (2021)[DOI]
- "2020Transportation Policy Evaluation Indicator Survey Project" (2021)[URL]
- C. Lee, "Crack Detection Technology of Lining Surface for Road Tunnel Maintenance Based on High Resolution Imaging Equipment and Artificial Intelligence", KSCE Journal of Civil Engineering, 69(9), pp.103~108 (2021)
- J. Kim, "Detection of Road Surface Defects Using Multiple Sensing Technologies", Journal of Korean Society for Nondestructive Testing, 43(1), pp.69~75 (2023)[DOI]
- E. Šabanovič, "Identification of Road-Surface Type Using Deep Neural Networks for Friction Coefficient Estimation", Sensors, 20(3), pp.612~629 (2020)[DOI]
- H. Yang, "Detection algorithm for road surface condition using wavelet packet transform and SVM", Proceeding of the 19th Korea-Japan joint workshop on Frontiers of Computer Vision, pp.323~326 (2013)
- D. Lee, "Intelligent Tire Sensor-Based Real-Time Road Surface Classification Using an Artificial Neural Network", Sensors, 21(9), pp.3233~3250 (2021)[DOI]
- W. Junyi, "Road surface real-time detection based on Raspberry Pi and recurrent neural networks", Transactions of the Institute of Measurement and Control, 43(11), pp.2540~2550 (2021)[DOI]
- "Final report of the national highway pavement management system" (2021)[URL]
- L. Cao, "Comparison of two deep learning model YOLOF and Detectron2 for mesoscale eddies identification in the South China Sea", International Journal of Remote Sensing, 45(19), pp.6919~6933 (2024)[DOI]
- M. Chung, "Modification of Deeplab V3+ for Detecting Cracks on Steel Rail Surfaces", Proceeding of International Conference on Consumer Electronics Taiwan, pp.325~326 (2024)