Large and complex steel structures play a vital role in building construction. However, deviations between the design model and the actual construction state are inevitable, which seriously affects the quality and safety of building construction. In our study, an intelligent construction monitoring method for large and complex steel structures based on laser point cloud is proposed. Firstly, three-dimensional laser scanning technology is introduced to capture accurate and complete spatial information on steel structures. Then, considering the inconsistency of the coordinate system between the design model and the laser point cloud, the building information model (BIM) is converted into the point cloud model, and the datum unification of the two types of the point cloud is achieved by adopting a coarse-to-fine registration strategy. Finally, the spatial information of steel structures is extracted from the laser point cloud based on the as-designed model, and the distance deviation between the two models is analyzed to reflect the actual construction state. To demonstrate the applicability of the proposed method, the steel structures’ point cloud of the stadium and the high-speed railway station is captured by the terrestrial three-dimensional laser scanner. The experimental results demonstrate that the method can extract the deviation between the design model and the actual construction, to provide accurate data sources for the intelligent fine construction of steel structures.

Spatially referenced and geometrically accurate laser scanning is essential for the safety monitoring of an underground mine. However, the spatial inconsistency of point clouds collected by heterogeneous platforms presents challenges in achieving seamless fusion. In our study, the terrestrial and handheld laser scanning (TLS and HLS) point cloud registration method based on the coarse-to-fine strategy is proposed. Firstly, the point features (e.g., target spheres) are extracted from TLS and HLS point clouds to provide the coarse transform parameters. Then, the fine registration algorithm based on identical area extraction and improved 3D normal distribution transform (3D-NDT) is adopted, which achieves the datum unification of the TLS and HLS point cloud. Finally, the roughness is calculated to downsample the fusion point cloud. The proposed method has been successfully tested on two cases (simulated and real coal mine point cloud). Experimental results showed that the registration accuracy of the TLS and HLS point cloud is 4.3 cm for the simulated mine, which demonstrates the method can capture accurate and complete spatial information about underground mines.

With the rapid development of autonomous driving technology and high-accuracy maps, accurate detection of lane line edges has become critical. Traditional methods, especially those relying on visual sensors, are often limited by light variations and environmental factors. Mobile laser scanning (MLS) systems is capable of generating high-accuracy 3D point cloud, but it still faces challenges such as incomplete information extraction and lack of precision in the process of extracting lane line edges. To solve these problems, this study proposes a highway lane line edge extraction method that combines an improved α-shape with a cubic B-spline curve fitting algorithm. Firstly, statistical filtering and radius filtering algorithms are used to remove isolated points from the point cloud and obtain pure highway laser point cloud. Based on the geometric properties and spatial proximity of the point cloud, adaptive voxel filtering is used for downsampling, and a region-growing algorithm is used to accurately extract the road point cloud. Secondly, based on the reflection intensity information of the point cloud, the optimal threshold is automatically determined with the help of the OTSU (Maximum Between-Class Variance Method) algorithm to separate the lane line point cloud from the road surface point cloud; and the shape parameters are adjusted based on the improved α-shape algorithm to control the completeness of the lane edge extraction, and to accurately extract the edge contour of the lane line. Finally, the improved cubic B-spline curve fitting method is used to approximate the generated lane edge fitting curves, and the vectorized extraction of the curves is achieved by introducing specific constraints. In this paper, point cloud experiments are conducted on both straight and curved sections of the highway to verify the reliability of the algorithm. Experimental results show that the root mean square error of lane line edge extraction and fitting is 7.8 mm in the curved section and 19 mm in the straight section, demonstrating that the proposed method is able to accurately extract the complete lane line edges of highways.