A Point Cloud Error Compensation Model Based on Additional System Parameters
S. Yue, D-J. Cai, D-J. Yue and C-Z. Zhai
The urban three-dimensional (3-D) model, which can truly and vividly express the 3-D spatial information of ground objects, was an essential link in the process of smart city construction. The backpack 3-D laser scanning systems had gradually become an important technical means to build urban 3-D models because of their high efficiency and high fineness. In this paper, the error source of backpack point cloud data was analysed based on the composition of backpack 3-D laser scanning system. The distribution of point cloud error was statistically analysed by designing experiments. It was found that there was a systematic error in the backpack 3-D laser scanning point cloud. There were many factors influencing the system errors and they were superimposed on each other. In order to solve this problem, an error compensation model with additional system parameters based on a multiple linear regression algorithm was constructed. The results showed that this model can weaken the influence of system errors on the backpack point cloud, effectively improve the accuracy of the backpack point cloud, and provide reliable point cloud data for subsequent processing of the building point cloud.
Keywords: Laser scanning, backpack, three-dimensional (3-D), point cloud, system errors, multiple linear regression