New Crack Detection Method for Bridge Inspection Using UAV Incorporating Image Processing

Abstract

Unmanned aerial vehicle (UAV) technologies combined with digital image processing have been applied to the crack inspection of bridge structures to overcome the drawbacks of manual visual inspection. However, because of environmental interference such as uneven natural light, noises produced by the UAV hardware, spots on the road surface, and UAV jitter, the collected images by UAVs are usually fuzzy and have relatively low contrast. In the processing of such collected images the traditional edge detection algorithms such as Canny algorithm, Prewitt algorithm, and Sobel algorithm have low detection accuracy because of their poor antinoise ability. K-means clustering method is one of the unsupervised learning methods. Nevertheless, in the case of a small amount of images, it cannot achieve the accurate identification of the cracks from the collected image. In this paper, a new crack detection method based on the crack central point, namely crack central point method (CCPM), is proposed to address these essential issues. With a small amount of images, the new method can quickly and accurately identify the cracks in the collected images. Compared with the traditional edge detection methods and K-means clustering method, the CCPM method has better adaptability and robustness.