资源简介

摘要： 为了提高激光点云滤波算法在地形复杂区域的精度、效率以及自适应性，基于布料模拟滤波算法，提出了一种面向大范围复杂场景的顾及地形特征的点云滤波方法。该方法首先采用基于坡度的动态格网分割方法，对剔除粗差后的点云建立格网索引；其次利用每个格网的邻域格网中的最低点建立曲面方程拟合高程值，通过计算真实高程与拟合高程差值实现高程归一化；然后使用布料模拟算法模拟布料下降过程得到地形布料的最终形态，进而通过阈值限定实现地面点提取。在地形复杂的测试区使用相同滤波参数进行算法改进前后对比测试，结果表明：改进算法的正确率由原CSF算法的88.9%提高到改进后算法的95.19%；I类误差、II类误差分别由9.71%、1.39%下降到4.57%、0.24%，且滤波时长由164 s缩减至60.9 s。本文提出的改进算法在保证大范围复杂场景区域滤波正确率的基础上，对不同地形具有较强的自适应性，且提高了滤波计算效率。 关键词： 机载激光点云;  点云滤波;  布料模拟滤波;  地形特征;  高程归一化     Abstract: In order to improve the precision, efficiency and robustness in complicated terrain conditions, a CSF-modified filtering method was presented based on Cloth Simulation Filtering(CSF). Firstly, with the method of dynamic grids based on slope, we established a gird index for the point cloud of which the gross errors are eliminated. Secondly, the terrain surface equation is produced by the nadir in every girds to fit elevation value of every point. And the difference between real and fitted value is the elevation normalized. Finally, with CSF algorithm to simulate gradually filtering process to refine the cloth and ground points are obtained by the final cloth and limited threshold. We verify the presented algorithm and the original CSF algorithm with the same filtering parameters in the complicated test area. The accuracy increases from 88.9% of original CSF algorithm to 95.19% of the CSF-modified algorithm. And error of type I and type II decrease from 9.71% and 1.39% to 4.57% and 0.24% respectively. In addition, the filtering duration is shortened from 164 s to 60.9 s. The result shows that on the basis of ensuring the accuracy of filtering in a wide range of complicated terrain, the modified method is not only adaptable for complicated situation areas but also helpful to enhance the efficiency of filtering computation.

相关资源

1. 水质遥感监测的关键要素叶绿素a的反演算法研究进展
2. 基于无人机遥感技术的森林参数获取研究进展
3. 基于无线传感器网络的森林生态系统观测试验平台构建
4. 基于无人机激光雷达的森林冠层高度分析
5. 多源时序国产卫星影像的森林火灾动态监测
6. 基于无人机遥感技术的台风灾害倒伏绿化树木检测

下载说明

·如果您发现该资源不能下载，请通知管理员.gissky@gmail.com

·为确保下载的资源能正常使用，请使用[WinRAR v3.8]或以上版本解压本站资源,缺省解压密码www.gissky.net ，如果是压缩文件为分卷多文件，请依次下载每一个文件，并按照顺序命名为1.rar,2.rar,3.rar...，然后鼠标右击1.rar解压.

·为了保证您快速的下载速度，我们推荐您使用[网际快车]等专业工具下载.

·站内提供的资源纯属学习交流之用,如侵犯您的版权请与我们联系.