基于三维激光点云的钢架螺栓球体节点要素提取研究
Research on Geometric Element Extraction of Steel Frame Bolts Based on 3D Laser Point Cloud

目前在新型机场、高铁站等复杂建筑的建设中，通常采用钢架结构构建主体框架，而此类钢架结构的内部施工需要获取螺栓球体的中心点坐标、直径等几何要素。为了高效、精确地提取螺栓球体的各几何要素，本文基于三维激光扫描技术获取点云数据，提出使用随机采样一致法对球体半径参数进行提取，通过随机采样一致法与最小二乘法拟合结果对比，发现随机采样一致算法对球体信息进行提取稳定性更好，精度更高。
At present, in the construction of large-span and complex buildings such as new airports and high-speed railway stations, the main frame is usually constructed by steel frame structure, and the internal construction of such steel frame structures needs to obtain the geometric elements of the center point coordinates and diameter of the bolt sphere. In order to extract the geometrical elements of the bolt sphere efficiently and accurately, this paper obtains the point cloud data based on the 3D laser scanning technology, and proposes to extract the parameters by using the random sampling consensus method. The results of the random sampling consistency method and the least squares fitting result are compared. The random sampling consensus algorithm extracts the sphere information better and has higher precision.