Adaptive up-sampling algorithm of point cloud model
点云模型适应性上采样算法

For insufficient sampling density of point cloud model, this paper proposed a new adaptive up-sampling algorithm. Firstly, it used uniform grids method to represent the spatial topology relationship of point cloud in order to improve the efficiency of finding the K-nearest neighbors for each data point, and estimated normal vectors of data points by constructing covariance matrix, and computed a consistent orientation of the normal vectors using normal propagation algorithm. Then, it detected these regions with insufficient sampling density dynamically, and adaptive resampled points uniformly in the tangent plane of bounding rectangle originated at this point of insufficient sampling density, and the re-samples were projected onto the underlying surface of point cloud model to achieve the final up-sampling result using the almost orthogonal projection. The up-sampling models could preferably supplement these little detail information of point cloud, and satisfy the needs of rendering point cloud model and subsequent geometric processing.