Automatic detection and classification of defect on road pavement using anisotropy measure

Existing systems for automated pavement defect detection can only identify cracking type defects. In this paper, we introduce a method which can detect not only cracks as small as 1mm in width, but also two other defect types: joint and bridged. Road images are captured by our acquisition system. Firstly, a pre-processing step is applied on images to remove lane-marking. Then an anisotropy measure is calculated to detect road defects. Finally, a backpropagation neural network is used to classify the images into four classes: defect-free, crack, joint and bridged. Experimental results were performed on real road images which were labelled by human operators. Comparisons with other methods are also given.