一种改进的渐进加密三角网点云滤波算法

An improved progressive TIN densification filtering algorithm for point clouds

针对渐进加密三角网滤波算法中初始不规则三角网对地形近似模拟的精度问题,该文提出了一种改进的渐进加密三角网滤波算法。该方法利用扩展局部极小值法,在包含点云的格网中提取待定种子点,并采用最邻近插值法获取不包含点云的格网高程;然后设定阈值来判断局部薄板样条插值的邻域高差,确定最终种子点;最后构建初始不规则三角网并迭代加密。结果表明,该方法能提高地面种子点提取的准确率,并进一步提高点云滤波的质量。

To solve the problem of the accuracy of approximating terrain simulation of initial irregular triangular network(TIN),an improved progressive TIN densification(IPTD) filtering algorithm was proposed. In this method,the undetermined seed points were selected by the method of extended local minimum for the grids containing points and the nearest neighbor method was used to interpolate the elevations of grids without points. Final seed points were determined by judging the elevation difference in the neighborhood of the local thin plate spline interpolation by the given threshold. Lastly,an initial triangular irregular network was constructed and encrypted iteratively. The result demonstrated that the algorithm was capable of improving the accuracy of extracting ground seed points and the quality of LiDAR point clouds filtering further.