2015 · 115 citations · 2 references
EngineeringCivil EngineeringComputer ScienceComputer VisionMachine VisionPavement EngineeringAirport PavementsImage AnalysisImage SegmentationAccuracyAdaptive Threshold SegmentationAirport Runway PavementsCrack DetectionCrack SegmentationPrecisionSegmentation-based DetectionThreshold-based Segmentation
In airport runway pavement crack detection, sign and marking on the runway pavement often disturb crack detection. They may even be mistaken for crack and result in lower accuracy. The crack detection based on twice-threshold segmentation technology is presented in this paper. Firstly, the road marking in the runway image are removed by using improved Otsu threshold segmentation algorithm. Secondly, the image after removed the marking is segmented by the improved adaptive iterative threshold segmentation algorithm to get the crack image. Finally, the image is denoise, and the final crack image is obtained. This method can effectively detect the crack in runway images containing road marking. The simulation experiments verify the effectiveness of the method and the algorithm has the higher precision.
2