Classification of surface pavement cracks as top-down, bottom-up, and cement-treated reflective cracking based on deep learning methods

The treatment and repair strategies for reflective and fatigue cracking that initiate at the pavement surface (i.e., top-down cracking) and at the bottom of the asphalt concrete layer (i.e., bottom-up cracking) are noticeably different. However, pavement management engineers are facing difficulties in identifying these cracks in the field because they usually appear in visually identical patterns. The objective of this study was to develop artificial neural network (ANN) and convolutional neural network (CNN) applications to differentiate and classify top-down, bottom-up, and cement-treated reflective cracking in in-service flexible pavements using deep-learning models. The developed CNN model achieved an accuracy of 93.8% in the testing and 91% in the validation phases, and the ANN model showed an overall accuracy of 92%. The ANN classification tool was developed based on variables related to pavement and crack characteristics including pavement age, annual average daily traffic, thickness of the asphalt concrete layer, type of base, crack orientation, and crack location.