梯度引导下的分块曲面LiDAR数据滤波方法

根据点云密度确定分块曲面滤波区块的大小,通过构建梯度引导图,引导进行分块曲面滤波,并利用得到的初始结果对滤波进行进一步细化。将最后滤波结果的总错误率与ISPRS提供的测试方法进行比对,在3个测区内总错误率均低于ISPRS提供的方法。实验结果表明,该方法不受建筑物大小的影响,在保证较好滤波结果的基础上,提高了算法的自适应性和滤波的自动化程度。

Block-index Curved Fitting Filtering Method Guidedby Gradient Based on LiDAR Data

LiDAR points clouds have become one of the most important data sources for 3.D buildingreconstruction.In this paper,the block size is determined by point cloud density.And the gradient between theblocks is used to guide the filtering process.For further refinement,the initial result is used to distinguish groundpoints and non-ground points.Results of the experiment show that this method can handle kinds of buildings.And on the basis of ensuring good filtering results,it makes the algorithm more adaptive and improves the degreeof automation.