空间分割与曲率相融合的点云精简算法研究

非接触式扫描方法获得点云数据存在大量的冗余数据。为便于模型重构,针对点云数据精简是必不可少的数据预处理手段,提出了一种基于空间分割和曲率特征信息的点云数据精简算法。通过K-邻域计算、二次曲面拟合、曲率估算和曲率阈值可调的数据分区等关键精简技术,实现了对同一数据不同区域应用不同精简算法,进行不同比例的数据精简。实例验证表明,该算法能适应各种类型曲面数据的精简要求,保证精简效率的同时,很好地保留点云的特征信息。

Study of point cloud data reduction algorithm integrating space partition and curvature

There are huge amounts of redundant data in point cloud data obtained by non-contact scanning. In order to realize model reconstruction effectively, point cloud data reduction is an indispensable means of pre-processing means. This paper presented an approach of point cloud data reduction based on space partition and curvature. Through some key technologies, such as K-neighborhood search, second surface fitting, curvature estimation, and data partition by of controllable curvature threshold, it applied the different reduction algorithms in different regions of the same point cloud data, meanwhile, achieved realizable reduction proportions. So, the algorithm can ensure reduction efficiency and retain characteristic information of the point cloud data simultaneously.