TY - GEN
T1 - Robust Road Surface Defects Detection Using Three-Dimensional Geometric Analysis and Adaptive Machine Learning Strategy
AU - Rehman, Zia Ur
AU - Ma, Hongbin
AU - Khan, Malak Abid Ali
AU - Wang, Xin
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.
PY - 2026
Y1 - 2026
N2 - Maintaining road surface integrity is vital for autonomous driving systems and vehicle safety. Traditional inspection methods are labor intensive, time consuming and subjective, highlighting the need for an efficient, automated process for the identification of road surface defects. In the proposed work, we present an effective method for identifying road surface defects using three-dimensional (3D) point cloud information. By employing 3D geometric analysis, such as surface normal and curvature’s calculation and apply an adaptive machine learning strategy, we accurately detect defects, even small minutes-level defects such as cracks and holes. Our approach combines geometric analysis with artificial intelligence (AI) through machine learning (ML) to classify defects based on point-cloud data, leading to a reliable and robust surface defects detection approach. This approach significantly improves the performance of the defect detector compared to the state-of-the-art detector on our custom dataset. This advancement promises to streamline road surface maintenance, ensuring the safety and reliability of humans and autonomous driving.
AB - Maintaining road surface integrity is vital for autonomous driving systems and vehicle safety. Traditional inspection methods are labor intensive, time consuming and subjective, highlighting the need for an efficient, automated process for the identification of road surface defects. In the proposed work, we present an effective method for identifying road surface defects using three-dimensional (3D) point cloud information. By employing 3D geometric analysis, such as surface normal and curvature’s calculation and apply an adaptive machine learning strategy, we accurately detect defects, even small minutes-level defects such as cracks and holes. Our approach combines geometric analysis with artificial intelligence (AI) through machine learning (ML) to classify defects based on point-cloud data, leading to a reliable and robust surface defects detection approach. This approach significantly improves the performance of the defect detector compared to the state-of-the-art detector on our custom dataset. This advancement promises to streamline road surface maintenance, ensuring the safety and reliability of humans and autonomous driving.
KW - Artificial Intelligence
KW - Defects Detection
KW - Machine Learning
KW - Three-Dimensional Geometric Analysis
UR - https://www.scopus.com/pages/publications/105028357449
U2 - 10.1007/978-3-032-13056-3_10
DO - 10.1007/978-3-032-13056-3_10
M3 - Conference contribution
AN - SCOPUS:105028357449
SN - 9783032130556
T3 - Communications in Computer and Information Science
SP - 116
EP - 129
BT - AI Revolution
A2 - Arabnia, Hamid R.
A2 - Deligiannidis, Leonidas
A2 - Amirian, Soheyla
A2 - Ghareh Mohammadi, Farid
A2 - Shenavarmasouleh, Farzan
PB - Springer Science and Business Media Deutschland GmbH
T2 - International Conference on the AI Revolution: Research, Ethics, and Society, AIR-RES 2025
Y2 - 14 April 2025 through 16 April 2025
ER -