Applicability analysis of cloth simulation filtering algorithm based on LiDAR data
-
摘要: 为了解决现有点云滤波算法设置参量多、滤波效果不理想、难以操作等问题，采用最新的布料模拟滤波算法，根据简单的物理过程构建虚拟格网模拟地形表面，并针对复杂地形的点云数据进行定性和定量滤波验证。结果表明，其Ⅰ类误差在5.7%、Ⅱ类误差在3.4%以内，但针对部分混合有平坦和陡坡的局部区域滤波效果并非理想；在满足滤波精度的同时，可在30s内完成数百万个点的滤波，甚至在数秒内完成数十万个点的滤波。该算法所需参量很少、效率非常高，能满足绝大多数复杂地形数据的滤波要求。Abstract: In order to solve the problems in the existing point cloud filtering algorithm, such as too many parameters, without ideal filtering effect and inconvenient operation, the cloth simulation filtering (CSF) algorithm was chosen to construct virtual grid to simulate terrain surface. The algorithm was verified using the qualitative and quantitative experiment and analysis in complex terrain point cloud data. The results show that type Ⅰ error is less than 5.7%, type Ⅱ error less than 3.4%, but for the local area with some mixed flat and steep slope the filtering effect is not ideal; The algorithm can achieve the filtering of millions of points in 30s while satisfying the filtering precision, and even hundreds of thousands of points in several seconds. The algorithm requires few parameters and is very efficient. It can satisfy the filtering requirements of most complex terrain data.
-
-
![]()
Figure 1. Overview of CSF algorithm
a—schematic illustration of CSF b—schematic illustration of the grid of CSF
![]()
Figure 2. Filtering results of data 1
a—original LiDAR point clouds b—LiDAR ground point after filtering c—digital terrain model d—terrain profile before and after filtering
![]()
Figure 3. Filtering results of data 2
a—original LiDAR point clouds b—LiDAR ground point after filtering c—digital terrain model d— terrain profile before and after filtering
![]()
Figure 4. Filtering results from data 3
a—original LiDAR point clouds b—LiDAR ground point after filtering c—digital terrain model d—point cloud filtering section
![]()
Figure 5. Filtering results of data 4
a—original LiDAR point clouds b—LiDAR ground point after filtering c—digital terrain model d— terrain profile before and after filtering
![]()
Figure 6. Filtering results of data 5
a—original LiDAR point clouds with elevation color b—reference point cloud after manual classification c—ground point clouds after filtering d—nonground point clouds after filtering
![]()
Figure 7. Filtering results of data 6
a—original LiDAR point clouds with elevation color b—reference point cloud after manual classification c—ground point clouds after filtering d—nonground point clouds after filtering
![]()
Figure 8. Statistics of time distribution
a—the running time of each step of the algorithm b—relationship between the running time of the algorithm and the numbers of the constructed grid nodes
Table 1 Characteristics of test data set of data 1~data 4(point)
dataset point number scope density features 1# 6185384 0.7km× 0.6km 11.81point/ m2 flat terrain, large buildings, potholes 2# 4978327 2.0km× 1.5km 1.61point/ m2 urban, buildings, vegetation 3# 23097984 0.6km× 0.9km 42.77point/ m2 mountain, steep slopes, dense vegetation 4# 19659989 1.2km× 0.5km 32.77point/ m2 mixture of flat terrain and hillside 
下载: 导出CSV
Table 2 Characteristics of test data set of data 5 and data 6
dataset point number density features ground point number not-ground point number 5# 418021 5.79point/m2 flat terrain，vegetation，building 252936 165085 6# 505619 6.57point/m2 vegetation on hillside，power line，low shrubs 439421 66198
下载: 导出CSV
Table 3 Two types of error statistics
category data 5 data 6 ground point number non-ground number ground point number non-ground number ground point number 252936 1658 439421 2251 non-ground number 4047 165085 25047 66198 typeⅠ 1.60% 5.7% typeⅡ 1.05% 3.4%
下载: 导出CSV
-
[1] TANG F F, RUAN Z M, LIU X. Research of filtering method for urban airborne LiDAR data[J]. Laser Technology, 2011, 35(4):527-530(in Chinese). http://en.cnki.com.cn/Article_en/CJFDTOTAL-JGJS201104023.htm
[2] SHAN J, APARAJITAN S. Urban DEM generation from raw lidar data:A labeling algorithm and its performance[J]. Photogrammetric Engineering & Remote Sensing, 2005, 71(2):217-226. http://cn.bing.com/academic/profile?id=317d6f374a0eee078ca4bc804b0fe309&encoded=0&v=paper_preview&mkt=zh-cn
[3] LIU C, CHEN H Y, WU H S. Data processing and feature extraction of LiDAR remote sensing[M]. Beijing:Science Press, 2010:189-195(in Chinese).
[4] XUELIAN M, NATE C, KAIGUANG Z. Ground filtering algorithms for airborne lidar data a review of critical issues[J]. Remote Sensing, 2010, 2(3):833-860. DOI: 10.3390/rs2030833
[5] ZHOU X M. Filtering algorithm of point cloud data for airborne LiDAR[D]. Zhengzhou: The PLA Information Engineering University, 2011: 1-4(in Chinese).
[6] AXELSSON P. Proeessing of laser scanner data-algorithms and applications[J]. ISPRS Journal of Photogrammetry & Remote Sensing, 1999, 54(2/3):138-147. http://www.sciencedirect.com/science/article/pii/S0924271699000088
[7] ZHANG X H. Theory and method of airborne lidar measurement technology[M]. Wuhan:Wuhan University Press, 2007:96-104(in Chinese).
[8] ZHANG X H. Airborne laser scanning altimetry data filtering and object extraction[D]. Wuhan: Wuhan University, 2002: 3-7(in Chinese).
[9] GEORGE S, GEORGE V. Experimental comparison of filter algorithms for bare-earth extraction from airborne laser scanning point clouds[J]. ISPRS Journal of Photogrammetry & Remote Sensing, 2004, 59(1/2):85-101. http://www.sciencedirect.com/science/article/pii/S0924271604000140
[10] MENG X, WANG L, SILVAN-CARDENAS J L. A multi-directional ground filtering algorithm for airborne LiDAR[J]. ISPRS Journal of Photogrammetry & Remote Sensing, 2009, 64(1):117-124. http://www.sciencedirect.com/science/article/pii/S0924271608000956
[11] SUSAKI J. Adaptive slope filtering of airborne LiDAR data in urban areas for digital terrain model (DTM) generation[J]. Remote Sensing, 2012, 4(6):1804-1819. DOI: 10.3390/rs4061804
[12] ZHANG K, CHEN S C, WHITMAN D. A progressive morphological filter for removing non-ground measurements from airborne LiDAR data[J]. IEEE Transactions on Geoscience & Remote Sensing, 2003, 41(4):872-882. http://ieeexplore.ieee.org/xpls/icp.jsp?arnumber=1202973
[13] MONGUS D, LUKAČ N, ŽALIK B. Ground and building extraction from LiDAR data based on differential morphological profiles and locally fitted surfaces[J]. ISPRS Journal of Photogrammetry & Remote Sensing, 2014, 93(7):145-156. http://www.sciencedirect.com/science/article/pii/S0924271613002840
[14] MONGUS D, ŽALIK B. Parameter-free ground filtering of LiDAR data for automatic DTM generation[J]. ISPRS Journal of Photogrammetry & Remote Sensing, 2012, 67(1):1-12. http://www.sciencedirect.com/science/article/pii/S0924271611001122
[15] LI Z L, ZHU Q. Digital elevation model[M]. Wuhan:Wuhan University Press, 2005:44-47(in Chinese).
[16] LAI X D. Fundamentals and applications of airborne LiDAR[M]. Beijing:Publishing House of Electronics Industry, 2010:150-201(in Chinese).
[17] ZHANG W M, QI J B, XIE D H. An easy-to-use airborne LiDAR data filtering method based on cloth simulation[J]. Remote Sensing, 2016, 8(6):501. DOI: 10.3390/rs8060501
[18] WEIL J. The synthesis of cloth objects[J]. ACM SIGGRAPH Computer Graphics, 1986, 20(4):49-54. DOI: 10.1145/15886
[19] PROVOT X. Deformation Constraints in a mass-spring model to describe rigid cloth behaviour[EB/OL].(2005-05-26)[2017-07-17]. http://www.cs.cornell.edu/courses/cs667/2005sp/studentSlides/07budsberg.pdf.
[20] MOSEGAARD J. Mosegaards cloth simulation coding tutorial[EB/OL].(2013-10-18)[2017-07-17]. http://victorfrench.com/cgi-techniques-clothsim.php.
计量
- 文章访问数: 19
- HTML全文浏览量: 1
- PDF下载量: 43