Double-filtering method for point cloud data in densely vegetated area
-
摘要: 为了解决目前机载激光雷达点云滤波算法中特征单一、运算效率低、植被覆盖区效果较差等问题，提出一种植被茂密地区的点云自适应双重滤波方法。首先利用回波分离方法，分别提取点云的单次回波和末次回波进行粗滤波处理；然后利用偏度平衡理论进行单次回波的强度阈值确定，同时利用最大类间方差法对首次回波和末次回波的高程差进行高差运算，实现末次回波高差阈值自动化，并融合粗滤波后单次回波和末次回波的点云数据；最后，利用不规则三角网渐进加密滤波算法对融合后的点云数据进行精滤波处理，并通过了实验验证。结果表明，3组数据集的Ⅱ类误差都相对较低，分别为1.06%, 1.64%, 1.34%。结合回波信息和高差信息的双重滤波方法不仅能较好地剔除植被，而且能较好地保留地形细节。Abstract: In order to solve the problems such as single feature, low calculation efficiency, and poor effect of vegetation coverage of the current airborne light detection and ranging(LiDAR) point cloud filtering algorithm, an adaptive double filtering method for point cloud in densely-vegetated areas was proposed. Firstly, the single and last echo of point cloud were extracted respectively and processed by coarse filtering by using the echo separation method. Then, the skewness balance theory was used to determine the intensity threshold of the single echo, and the maximum inter-class variance method was used to calculate the height difference between the first echo and the last echo, so as to realize the automation of the height difference threshold of the last echo. And the point cloud data of the single echo and the last echo after rough filtering were integrated. Finally, the incremental encryption filtering algorithm of triangulated irregular network was used to carry out the fine filtering processing on the fused point cloud data, and the experimental verification had been done. The results show that the type Ⅱ errors of the three data sets are relatively low of 1.06%, 1.64%, and 1.34% respectively. The dual filtering method that combines echo information and height difference information can not only eliminate vegetation, but also retain terrain details.
-
-
![]()
Figure 4. Experimental data single echo result graph
a—data set 1 b—data set 2 c—data set 3
![]()
Figure 5. Figure of final echo results of experimental data
a—data set 1 b—data set 2 c—data set 3
![]()
Figure 6. Figure of coarse filtering result of point cloud
a—data set 1 b—data set 2 c—data set 3
![]()
Figure 7. Figure of experimental results of point cloud precision filtering
a—data set 1 b—data set 2 c—data set 3
Table 1 Precision evaluation
sample filtered data total ground point non-ground point ground point a b e=a+b non-ground point c d f=c+d total p=a+c q=b+d h=a+b+c+d 
下载: 导出CSV
Table 2 Error evaluation table of experimental results
error data set 1 data set 2 data set 3 type Ⅰ/% 2.12 3.23 3.09 type Ⅱ/% 1.06 1.64 1.34 total/% 1.64 2.10 1.89
下载: 导出CSV
-
[1] HUI Z Y, CHENG P G, GUAN Y L, et al. Review on airborne LiDAR point cloud filtering[J]. Laser & Optoelectronics Progress, 2018, 55(6): 060001(in Chinese).
[2] YU Y Y. Research on filtering and classification algorithms of LiDAR data[D]. Hefei: University of Science and Technology of China, 2020: 19-20(in Chinese).
[3] CHEN C F, GAO Y, LI Y. A feature-preserving point cloud denoising algorithm for LiDAR-derived DEM construction[J]. Survey Review, 2021, 53(377): 146-157. DOI: 10.1080/00396265.2019.1704562
[4] MOUDRY V, KLAPSTĚ P, FOGL M, et al. Assessment of LiDAR ground filtering algorithms for determining ground surface of non-natural terrain overgrown with forest and steppe vegetation[J]. Measurement, 2020, 150: 107047. DOI: 10.1016/j.measurement.2019.107047
[5] LIU Y, SUN Sh Y. Laser point cloud denoising based on principal component analysis and surface fitting[J]. Laser Technology, 2020, 44(4): 497-502(in Chinese).
[6] WANG M H, ZHANG X H, ZENG T, et al. Preprocessing algorithms for filtering airborne LiDAR data[J]. Geomatics and Information Science of Wuhan University, 2010, 35(02): 224-227(in Chinese).
[7] LIU J N, ZHANG X H. Classification of laser scanning altimetry data using laser intensity[J]. Geomatics and Information Science of Wuhan University, 2005, 30(3): 189-193(in Chinese).
[8] ZHANG X H. A irborne LiDAR points cloud classification based on difference of twice return pulse heights[J]. Science of Surveying and Mapping, 2006, 31(4): 48-50(in Chinese).
[9] TANG F F, LIU J N, ZHANG X H, et al. A voxel-based filtering algorithm for DTM data extraction in forest areas [J]. Journal of Beijing Forestry University, 2009, 31(1): 55-59(in Chinese).
[10] TANG F F, RUAN Zh M, LIU X. Research of filtering method for urban airborne LiDAR data[J]. Laser Technology, 2011, 35(4): 527-530(in Chinese).
[11] LIN X G, ZHANG J X. Analysis of multiple returns of airborne LiDAR pulse and its application in filtering[J]. Science of Surveying and Mapping, 2013, 38(2): 22-25(in Chinese).
[12] SUN D, LI Y Q, SUN Zh Zh, et al. Research on the point cloud filter-ing for steep slope based on multi-pulse and fisher discrimination algorithm[J]. Geography and Geo-Information Science, 2018, 34(2): 34-39(in Chinese).
[13] LI P T, ZHAO Q Zh, CHEN H. Filtering UAV LiDAR point cloud with k-means clustering under the constraint of echo intensity[J]. Journal of Geo-information Science, 2018, 20(4): 471-479(in Ch-inese).
[14] BARTELS M, WEI H. Segmentation of LiDAR data using measures of distribution[J]. International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences, 2006, 36(7): 31-426.
[15] CROSILLA F, MACORIG D, SCAIONI M, et al. LiDAR data filter-ing and classification by skewness and kurtosis iterative analysis of multiple point cloud data categories[J]. Applied Geomatics, 2013, 5(3): 225-240. DOI: 10.1007/s12518-013-0113-9
[16] OTSU N. A threshold selection method from gray-level histograms[J]. IEEE Transactions on Systems Man & Cybernetics, 1979, 9(1): 62-66.
[17] DONG Q, WANG J Q, SUN Y G. An improved two-dimensional OTSU segmentation method for nvshu character image[J]. Journal of Multimedia, 2014, 9(9): 1081-1088.
[18] AXELSSON P. DEM generation from laser scanner data using adaptive TIN models[J]. International Archives of Photogrammetry Remote Sensing, 2000, 33(4): 110-117.
[19] LIU Y, XI X H, WANG Ch, et al. An improved progressive TIN densification filtering algorithm for point clouds[J]. Science of Surveying and Mapping, 2020, 45(5): 106-111(in Chinese).
[20] XING Ch B, DENG X Sh, LI Y X. Seed point optimal selection method of PTD algorithm based on confidence interval estimation th-eory[J]. Engineering of Surveying and Mapping, 2020, 29(5): 27-32(in Chinese).
[21] TIAN W W. A study on data filtering and dem extraction of lidar point cloud[D]. Changchun: Changchun University of Science and Technology, 2018: 23-24(in Chinese).
-
期刊类型引用(7)
1. 胡泽雄，游利兵，寸超，王宏伟，范军，王琪，张艳琳，方晓东. 准分子激光低抖动延时同步系统. 量子电子学报. 2023(01): 69-78 . 
百度学术
2. 龚玮玮，张进，张丽伟. 基于嵌入式技术的准分子激光器智能控制系统. 激光杂志. 2021(10): 181-185 . 百度学术
3. 王晨，梁勖，林颖，方晓东. MOPA结构准分子激光同步触发设计. 红外与激光工程. 2021(11): 165-170 . 百度学术
4. 朱志坚，薛竣文，王玉珂，孙鲁，苏秉华. 基于MOPA结构的1064nm单频光纤激光器. 激光技术. 2019(06): 800-803 . 本站查看
5. 王景景，符志军. 激光陀螺仪的机械抖动控制技术分析和研究. 激光杂志. 2019(12): 118-122 . 百度学术
6. 谢正兰，万川梅. 单向链式网络的激光同步技术研究. 激光杂志. 2018(03): 146-150 . 百度学术
7. 徐学红，栗科峰. 激光接收器的抖动抑制优化控制方法. 激光杂志. 2017(06): 136-139 . 百度学术
其他类型引用(2)
计量
- 文章访问数: 6
- HTML全文浏览量: 0
- PDF下载量: 7
- 被引次数: 9