A potential crack region method to detect crack using image processing of multiple thresholding

Abstract In this paper, a potential crack region method is proposed to detect road pavement cracks by using the adaptive threshold. To reduce noises of image, pre-treatment algorithm was applied according following steps: grayscale processing, histogram equalization, filtering traffic lane. From image segmentation methods, combines global threshold and local segment image. According distribution characteristics sliding window used obtain deviation, then, deviation segmented based on maximum inter-class deviation. Obtain then perform threshold-based algorithm. Real images surface were at Su Tong Li in Suzhou, China. It found that approach could give more explicit description images. The tested 509 German asphalt distress (Gap) dataset: test results be promising (precision = 0.82, recall 0.81, F 1 score 0.83).