| description abstract | In order to study the precise description and feature quantification of the tunnel lining crack morphology, the paper uses the tunnel lining panoramic-image method to rely on the Wujing Expressway Qingyantou Tunnel as the project, and the lining crack image acquired by the tunnel lining expansion image generator to carry out the feature points. Extraction, matching, model recognition, image expansion, and splicing processing are used to obtain a comprehensive reflection of the tunnel lining panoramic expansion image. The algorithm is analyzed from the aspects of image enhancement preprocessing, image edge detection, crack feature area interference point removal, crack connection, and feature statistics, and the boundary position of the crack is found by the connected area labeling method to determine the spatial position and angular direction of the defect in the image. Using the principle of crack pixel position relationship to calculate the crack length, so as to extract the detailed feature information of the lining crack, a tunnel surface lining crack identification evaluation system was developed, which provides a convenient, efficient, and comprehensive process for tunnel lining crack detection. | |
