Abstract: In order to solve the problems of traditional single-point measurement techniques based on total stations such as difficulty in obtaining overall information of steel structures, slow measurement speed and low model reconstruction accuracy, this paper applies the three-dimensional laser scanning technology to the field of steel structure model reconstruction, and proposes a complex steel structure model reconstruction method based on LiDAR point cloud. Firstly, the high-precision registration of multi-station point cloud is realized based on the target sphere and structural plane information, and the noise is removed by radius filtering algorithm to obtain the complete laser point cloud data. Then,the single model is segmented to realize the model reconstruction of each component of steel structure. Finally, the accuracy of the reconstructed three-dimensional model is evaluated according to the field measured data. The experimental results show that the accuracy of the model obtained by this method is high, and the difference between the model and the actual size is within 1 cm, which provides an effective solution for the three-dimensional model reconstruction of complex steel structure.

Key words: LiDAR point cloud, complex steel structure, radius filtering, model reconstruction