This study proposed a LiDAR-based approach for mapping water-related structures through point cloud registration.Point cloud data was collected from river sections and sluices.Appropriate source and target point clouds were selected,and coarse registration was performed based on coordinate transformation.Precise registration was carried out using an optimized iterative closest point algorithm.The optimal transformation matrix was solved using the least squares method.These steps enabled the achievement of data transformation and point cloud registration.The experimental results showed that the registration method in this paper could improve the point cloud information and realize the complementarity of point cloud information.In the number of point clouds,the registered data increased by 39 807 relatives to the target point cloud;in the detection distance and the relative target point cloud was expanded by 20.87 m.Compared with the traditional iterative closest point and normal distributions transform algorithms,the registration time was reduced by 31.63%and 3.21%,respectively.In terms of registration accuracy,the error was 4.30%higher than the iterative closest point algorithm and 21.08%lower than the normal distributions transform algorithm.
intelligent water conservancyLiDARcoordinate transformationleast-square methoditerative closest point
维普
万方数据
