Transportation Research Record Journal of the Transportation Research Board · 1995 · 24 citations · 0 references
EngineeringCivil EngineeringComputer ScienceComputer VisionMachine LearningPavement EngineeringCrack AssessmentImage ClassificationPattern RecognitionPavement DistressesPavement EvaluationPavement MaterialsCracksDistress AssessmentFuzzy SetsPavement Distress AnalysisPavement Distress AssessmentPavement Distress EvaluationPavement Distress IdentificationPavement Distress Quantification
A novel approach of applying the theory of fuzzy sets and moment invariants to analyze pavement images in proposed in this paper. By applying the theory of fuzzy sets and calculating moment invariants from different types of distress, features are obtained. Then, a back-propagation neural network is used to classify these features. The crack density is used to obtain extent information. This approach is illustrated using randomly selected samples from NCHRP Project 1-27 video images of real cracks. Based on these samples, the feasibility of using the theory of fuzzy sets and moment invariants to classify different types of crack is proven. High accuracy of classification is also obtained.