Objectives Point cloud filtering is the basic work of extracting ground information, generating digital elevation model and other terrain products. The slope filtering algorithm for point cloud is simple and easy to implement. In order to further improve the adaptability of slope filtering algorithm, a multi-scale adaptive slope filtering algorithm for point cloud is proposed.
Methods The idea of multi-scale pseudo-grid is used for reference, and the filtering threshold of grid is determined by combining k-means clustering and normal distribution adaptive to realize multi-scale adaptive point cloud slope filtering. Firstly, a pseudo-grid is introduced to divide the point cloud on the basis of data pre-processing. Then, the slope angle of grid points is calculated one by one using distance weighting, and the threshold value is determined adaptively by combining k-means clustering and normal distribution. Finally, the multi-scale strategy is used to reduce the grid size step by step to achieve fine filtering of point cloud data.
Results The algorithm is verified using two point cloud data sets with different densities. Compared with the two slope filtering algorithms and the classical algorithms of International Society for Photogrammetry and Remote Sensing, the experimental results show that this algorithm has better overall fitting results and higher stability.
Conclusions Through the above experiments, the proposed method can be applied to the filtering of point cloud data in different scenes.
DownLoad: