Detection of crack on concrete surfaces using image processing techniques

Abstract

Cracks on the concrete surface are one of the earliest indications of structural disintegration under maintenance; further the continual exposure can cause severe harm to the environment. The manual method of crack detection is time-consuming. Since the manual approach fully depends on the specialist’s data and experience, it lacks perspicacity within quantitative analysis. This study establishes a MATLAB® programme on image process techniques for automatic crack recognition and analyses. An image pre-processing theme combines multiple adaptive filtering and contrast enhancement based on the image that may improve the removal of background noise. Then, for removing isolated noise spots, it tends to develop a local adaptive threshold segmentation formula combined with an image morphological operator. The goal of this Gray intensity adjustment methodology is to improve the accuracy of the crack detection results. Furthermore, the results show that the target crack is identified on the concrete surface, additionally classified as its crack type and its properties are extracted.