UAV Dam Crack Detection System Based on Beidou and LIDAR

With the rapid development of dam construction, the demand for dam management and maintenance is rapidly expanding. Dam cracks are the core of dam management and maintenance, and rapid and effective detection of cracks and their spatial location and geometric parameters are the basis for dam maintenance. However, dam safety supervision mainly relies on manual work, which is very labor-intensive and time-consuming, and the detection is not fine enough. Therefore, we propose a new and efficient low-cost dam inspection system - UAV dam crack detection system based on Beidou and LIDAR. The UAV is equipped with LIDAR to obtain a high-precision three-dimensional point cloud of the dam surface, use the filtering algorithm to obtain the height difference and intensity contrast of the same dam surface, and then obtain the candidate point cloud of cracks according to the height difference and intensity contrast, use the maximum entropy threshold segmentation method to extract the candidate point cloud, and then filter and denoise the candidate point cloud based on the trough effect, and extract the cracks and their geometric parameters according to the morphological filtering. The texture information obtained from the conventional optical camera is used to construct a 3D model of the dam, and the cracks are displayed on the 3D model to statistically evaluate the severity of the cracks. The system aims to get more and more accurate information about the location and geometry of dam cracks through the combination of Beidou system and LIDAR, so as to better detect the hidden danger of dam safety, save human and material resources for dam safety management, and make the management more convenient and efficient.