Abstract
To quickly obtain bridge surface crack information to assist bridge condition assessment or maintenance, computer vision technology is used in this paper to quantitatively identify bridge surface cracks. The features of the bridge surface in different parts of the bridge may be quite different. To carry out crack identification research in a targeted manner, first, a classification model is trained based on deep learning to discriminate the surface pattern type of bridges. Then, for the images of bridge surfaces with different features, classification and segmentation models are used to identify the location and profile of the crack in the image; in other words, the existence of the crack is first judged by the trained classification models and pixel-level segmentation is performed by the trained segmentation models on the local region where the crack exists to determine the profile of the crack. Finally, based on the trained models of three types, crack identification is realized in combination with the improved sliding window in high-resolution bridge surface images with multiclass features. According to the identification results, the unconnected cracks are separated by the connected domain. From the results, a general algorithm for calculating the length and width of each crack is established, realizing the quantification of crack identification results regardless of the surface crack form.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.