Lidar based edge-detection for bridge defect identification

Previous visual damage detection on bridge structure based on eye-ball method is arbitrary and time-consuming for bridge management due to its heuristic nature. Commercial remote sensing (CRS), which has remarkable applications for geometric quantification, is suggested to supplement visual bridge inspection. Ground-based LIDAR is one of the remote sensing tools that have been successfully used in bridge evaluation. Most of the early measurement algorithms are developed based on the spatial information contained from the LIDAR data; this paper explores the potential of applying another important feature of the scan data: the reflectivity, to enhance the defect detection program. The addition of reflectivity in damage diagnostics is particularly useful for defect detection of curved surfaces. A damaged joint area and concrete beam were selected to verify the method. The study shows that the reflectivity of the LIDAR could be used to support the automatic defect detection in bridges by combining it with the current position-based only image processing algorithms.