Novel approach to pavement cracking detection based on morphology and mutual information

The image processing technology has been widely applied to pavement cracking detection, but it is difficult to make a balance between the processing effect and speed. In order to solve this problem, the multi-direction structuring elements morphology and mutual information algorithm for the pavement crack detection is introduced and a novel image processing method is proposed in this paper. The approached method combines the multi-direction structuring elements morphology with mutual information to segment the crack and the background. After the above processing, the proposed method employs the improved contour tracking algorithm to mark the cracks and obtain the contours, and then gets the length and width of cracks by counting. Compared with other traditional methods, the experimental results indicate that our method can not only effectively extract the edges of cracks and evidently decrease the influence of noises, but also gain the exact characteristics of the pavement cracks. Moreover, it can provide good image processing speed.