Raveling is one of the most common asphalt pavement distresses that occur on U.S. highway pavements. A raveling condition survey is required for highway agencies to determine the severity levels, the extents, and the locations of raveling so the preservation or rehabilitation treatments can be appropriately applied. However, the traditional raveling survey method, including determination of the raveling severity level (e.g. Low, Moderate, or High; or Level 1, 2, or 3), extent, and location is a visual inspection that is very time consuming, subjective, and hazardous to highway workers. In this project, we develop successful and effective raveling detection, classification, and measurement algorithms using 3D pavement data and macro-texture analysis, and to comprehensively validate these methods using large-scale, real-world data. The developed algorithms have been tested and validated using the pavement condition survey protocol in the Georgia Department of Transportation (GDOT). The algorithms can be extended to other highway agencies’ pavement condition survey protocols by re-training the classification components using corresponding ground truth data.
Raveling Example
GIS Visualization of Raveling Classification Result on Interstate Highway 285, Atlanta
Qualitative Evaluation of Level-1 Raveling Classification Results on I-285
Automatically Detect and Classify Asphalt Pavement Raveling Severity Using 3D Technology and Machine Learning,
Yi-Chang (James) Tsai, Yipu Zhao, Bruno Pop-Stefanov, Anirban Chatterjee, International Journal of Pavement Research and Technology, 2020 (paper).