A Framework for Automated Pavement Condition Monitoring

Abstract

Pavement condition monitoring is mainly performed manually. Inspectors are driving or walking the road network bare eyed to look for irregularities. Moreover, processing the collected data for understanding the road condition is also a manual task. In this paper a framework that automates the process is presented. Video data collected from the car's parking camera is utilized to detect defects in frames. Simultaneously, elevation signals collected from accelerometers attached to the car are processed to reconstruct the profile of the road and detect defects associated with its z-axis, such as bumps. A GPS device is synchronized with the other sensors to acquire the data's geolocation. Detected defects are then classified according to their type and their severity is assessed. All information is then transferred via 4G network to a central server, where the Road Condition Index of road segments necessary to classify roads is calculated. Finally, everything is saved in a Pavement Management System. Preliminary results on the processing of video data demonstrate the frameworks' promising application. The initial identification of frames including defects produces an accuracy of 96% and approximately 97% precision. Further experiments on such frames, aiming at the detection of potholes, patches and three different types of cracks result in over 84% overall accuracy and over 85% precision.

Computer Vision‐Based Patched and Unpatched Pothole Classification Using Machine Learning Approach Optimized by Forensic‐Based Investigation Metaheuristic

“…Ryu et al [14] attempt to improve an existing intelligent transportation system service by the utilization of a 2D image-based pothole detection with image processing techniques of segmentation, region extraction, and morphological filtering. A novel approach based on semantic texton forests used with 2D video frames has been proposed in [15]. Kamaliardakani et al [16] employ a heuristic thresholding method for detecting sealed crack damages.…”

Pemanfaatan Teknologi UAV (Unmanned Aerial Vehicle) Untuk Identifikasi dan Klasifikasi Jenis-jenis Kerusakan Jalan

“…Penelitian yang dilakukan oleh [13] menemukan bahwa data kerusakan jalan dapat dikumpulkan melalui rekaman video menggunakan kamera yang dipasang pada mobil. Elevasi permukaan yang direkam dari akselerometer pada mobil diproses untuk merekonstruksi profil melintang serta kedalaman lubang pada jalan.…”

Monitoring of Concrete Placement and Vibration for Real-Time Quality Control

“…Zhu et al [9] introduced a method that utilizes image processing techniques for automated detection and measurement of surface air pockets on concrete. In the study of Radopoulou et al [10], a framework that automates the process of acquiring data of defects on a road such as cracks, patches, and potholes was proposed. In addition to the inspection tools, there have been efforts in developing and improving defect management information systems to encourage utilization of detect data by streamlining data input, storage, and retrieval.…”