Automatic Asphalt pavement crack detection and classification using Neural Networks

Managing of road maintenance is the most complex task for road administrations. The first presumption for the evaluation analysis and correct road construction rehabilitation is to have accurate and up-todate information about road pavement condition. As the pavement condition survey is a critical process, it needs fast and cost-effective methods to collect necessary data. The paper proposes a system for automatic road pavement survey that uses image processing techniques to extract features from road images. A Neural Networks approach is used for detection of regions of images with defects and, further processing also, classifying defects into separate types. Proposed system could be used in the future to replace human labour for identification and classification of defects.