Research and Implementation for Scattered Point Cloud Data Denoising Method

In the science and technology of Surveying and Mapping, 3D laser scanning technology is an important means of Surveying and mapping and geographic information data acquisition. But there is noise and redundant data in the original point cloud data collecting by 3D laser scanning technology. The paper analyzed systematically point cloud data of the different types, the noise causes and consequence, and according to the categories summarized various denoising methods. On the basis of studying the law of scattered point cloud data, the denoising algorithms were analyzed to Laplacian, Mean Curvature Flow filtering, Mean Shift, Bilateral Filtering and so on. And their denoising ideas were illuminated, the program has implemented algorithm by Matlab software compilation, the feasibility of these denoising algorithms were verified by the cases. The research results provide a reference for point cloud data optimization, and create a good foundation for the point cloud data processing and 3D model construction, which is better serve the construction of digital city, smart city and other fields.