A novel road crack detection and identification method using digital image processing techniques

• Published in: 2012 8th International Conference on Computing and Convergence Technology pp. 397 - 400
• Main Authors: Weiling Huang, Ning Zhang
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.12.2012
• Subjects: crack detection; gap connect; Image preprocessing; Image segmentation; Lighting; Maintenance; mathematical morphology; Morphology; Neural networks; Noise; Object segmentation; road crack; Roads; Thresholding (Imaging)
• ISBN: 1467308943; 9781467308946

Road crack detection is the base of highway maintenance and digital image processing has been widely used in crack detection and identification. In this paper, a novel, effective, efficient image processing method is proposed for the extraction of road cracks from pavement images. The first stage is image preprocessing which consists of binary segmentation, morphological operations and remove algorithm which remove the isolate dots and area. Usually, after those operations above, many gaps still exists in the crack, the second stage proposed a novel algorithm to connect those break cracks. It needs to judge the type of the crack because of the difference in different types. The experimental shows that the surface crack could be detected correctly and results proved satisfying and efficient.