基于区域特征分割的密集匹配点云渐进形态学滤波

doi:10.12082/dqxxkx.2018.0452

摘要/Abstract

摘要：

随着计算机视觉和遥感技术的进步,基于遥感影像的密集匹配也成为目前获取高精度点云的重要手段之一。与LiDAR点云类似,点云数据处理的基础步骤就是点云滤波。在数据特征上,密集匹配生成的点云与LiDAR获取的点云既类似但又有区别。本文在渐进形态学滤波算法上添加了特征条件,将点云和图像结合成深度图像,并对深度图像按典型地物类型进行语义分割,从而对与图像平面坐标一致的点云进行标记和首次滤波;然后按几何特征将场景简单分类,按分类结果对应的参数滤波构建地面点三角网;最后综合初滤波结果和语义分割类型标记对特征相似的区域进行优化确认,得到最终的滤波结果,并与布料模拟滤波（CSF）算法进行了对比验证实验。结果表明,基于特征的渐进形态学滤波其I类误差在1.98%以内,Ⅱ类误差在2.33%以内,较适宜对精度要求较高的应用,尤其是混合地形的滤波。

关键词: 密集匹配, 点云滤波, 布料滤波, 深度学习, 区域特征分割, 渐近形态学滤波, 无人机

Abstract:

With the progress of computer vision and RS, dense matching based on remote sensing images has also become one of the important means to obtain high-precision point clouds. Like point clouds of LiDAR, filtering is the fundamental step. Dense matching point cloud is similar with LiDAR point cloud, but have different feature. In this paper, the feature condition is added to the progressive morphological filtering algorithm, point clouds and images are combined into RGB-Depth images, and depth images are semantically segmented according to typical object types, so that point clouds which coordinate correspond with image coordinate are marked and filtered for the first time. Then divide point clouds by grid, then do simply classified according to geometric features, and the improved irregular triangular network of ground points is constructed by filter parameters corresponding to the classification results. Finally, use and intergraded the pre-filtering results and the semantic segmentation results, the regions with similar features are optimized and confirmed by predefined parameter, and the final filtering results are obtained. The results are compared with results of the Cloth Simulated Filtering algorithm. The test result was show that type I error less than 1.98%, type II error less than 2.33% of the progressive morphological filtering algorithm, that algorithm is suitable for higher precision application, especially mixed terrain points cloud filtering.

Key words: dense matching, point cloud filtering, cloth simulation, deep learning, region feature segmentation, morphological filtering, UAV

张刚, 刘文彬, 张男. 基于区域特征分割的密集匹配点云渐进形态学滤波[J]. 地球信息科学学报, 2019, 21(4): 615-622.DOI:10.12082/dqxxkx.2018.0452

Gang ZHANG, Wenbin LIU, Nan ZHANG. Progressive Morphological Filtering Method of Dense Matching Point Cloud based on Region Feature Segmentation[J]. Journal of Geo-information Science, 2019, 21(4): 615-622.DOI:10.12082/dqxxkx.2018.0452

图/表 6

图1

图2

图3

表1

图4

表2

参考文献 23

[1]	惠振阳,程朋根,官云兰,等.机载LiDAR点云滤波综述[J].激光与光电子学进展,2018(6):1-9.
	[ Hui Z Y, Cheng P G, Guan Y L, et al.Review on airborne LiDAR point cloud filtering[J]. Laser & Optoelectronics Progress, 2018(6):1-9. ]
[2]	廖小罕,周成虎,苏奋振,等.无人机遥感众创时代[J].地球信息科学学报,2016,18(11):1439-1447.
	[ Liao X H, Zhou C H, Su F Z, et al.The mass innovation era of UAV remote sensing[J]. Journal of Geo-information Science, 2016,18(11):1439-1448. ]
[3]	时培强,江虹.机载LiDAR点云数据的多级滤波方法[J].通信技术,2018(1):67-74.
	[ Shi P Q, Jiang H.Multilevel method for airborne LiDAR point clouds data filtering[J]. Communications Technology, 2018(1):67-74. ]
[4]	陈香,王琳,张晔.基于SIFT的无人机遥感影像匹配算法研究[J].测绘与空间地理信息,2013(4):106-108. doi: 10.3969/j.issn.1672-5867.2013.04.033
	[ Chen X, Wang L, Zhang Y.The research on matching algorithm of the unmanned aerial vehicle remote sensing images based on SIFT[J]. Geomatics & Spatial Information Technology, 2013(4):106-108. ] doi: 10.3969/j.issn.1672-5867.2013.04.033
[5]	张彦峰. 利用多条件约束的航空影像逐像素密集匹配算法研究[D].北京:中国测绘科学研究院,2014.
	[ Zhang Y F.Pixel-wise dense matching of aerial images with multi-conditional constraints[D]. Bei[J].ng: Chinese Academy of Surveying And Mapping, 2014. ]
[6]	邹小丹. 基于半全局优化的多视影像匹配方法与应用[D].长沙:中南大学,2013.
	[ Zhou X D.A multi-view image matching method based on semi-global optimization and its application[D]. Changsha: Central South University, 2013. ]
[7]	Hirschmuller H.Accurate and efficient stereo processing by semi-global matching and mutual information[C].IEEE Conference on CVPR.San Diego, 2005:807-814.
[8]	杨化超,姚国标,王永波.基于SIFT的宽基线立体影像密集匹配[J].测绘学报,2011,40(5):537-543.
	[ Yang H C, Yao G B, Wang Y B.Dense matching for wide base-line stereo images based on SIFT[J]. Acta Geodaetica et Cartographica Sinica, 2011,40(5):537-543. ]
[9]	张力,张继贤.基于多基线影像匹配的高分辨率遥感影像DEM的自动生成[J].测绘科学,2008(52):35-39.
	[ Zhang L, Zhang J X.Multi-image matching for DEM generation from satellite imagery[J]. Science of Surveying and Mapping, 2008(52):35-39. ]
[10]	吴军,姚泽鑫,程门门.融合SIFT与SGM的倾斜航空影像密集匹配[J].遥感学报,2015,19(3):431-442.
	[ Wu J, Yao Z X, Cheng M M.Airborne oblique stereo image dense matching by integrating SIFT and SGM algorithm[J]. Journal of Remote Sensing, 2015,19(3):431-442. ]
[11]	闸旋,王慧,程挺,等.基于GPGPU的数学形态学LiDAR点云快速滤波方法[J].测绘科学技术学报,2013,30(1):73-77. doi: 10.3969/j.issn.1673-6338.2013.01.017
	[ Zha X, Wang H, Cheng T, et al.Morphological LiDAR points cloud filtering method based on GPGPU[J]. Journal of Geomatics Science and Technology, 2013,30(1):73-77. ] doi: 10.3969/j.issn.1673-6338.2013.01.017
[12]	Zhang W, Qi J, Wan P, et al.An easy-to-use airborne LiDAR data filtering method based on cloth simulation[J]. Remote Sensing, 2016,8(6):501. doi: 10.3390/rs8060501
[13]	Zhang W, Chen Y, Wang H, et al.Efficient registration of terrestrial LiDAR scans using a coarse-to-fine strategy for forestry applications[J]. Agricultural and Forest Meteorology, 2016,225:8-23.
[14]	张昌赛,刘正军,杨树文,等.基于LiDAR数据的布料模拟滤波算法的适用性分析[J].激光技术,2018(5):7-15.
	[ Zhang C S, Liu Z J, Yang S W, et al.Applicability analysis of cloth simulation filtering algorithm based on LiDAR data[J]. Laser Technology,2018(5):7-15. ]
[15]	李峰,崔希民,袁德宝,等.改进坡度的LiDAR点云形态学滤波算法[J].大地测量与地球动力学,2012,32(5):128-132.
	[ Li F, Cui X M, Yuan D B, et al.Slope improved morphological filtering algorithm for LiDAR point clouds[J]. Journal of Geodesy and Geodynamics, 2012,32(5):128-132. ]
[16]	梁鑫,杨晓云.基于坡度自适应的机载LiDAR分割算法[J].测绘科学,2013(2):72-74.
	[ Liang X, Yang X Y.Slope adaptive based segmentation for airborne LiDAR[J]. Science of Surveying and Mapping, 2013(2):72-74. ]
[17]	惠振阳,胡友健.基于LiDAR数字高程模型构建的数学形态学滤波方法综述[J].激光与光电子学进展,2016(8):7-13.
	[ Hui Z Y, Hu Y J.Review on morphological filtering algorithms based on LiDAR digital elevation model construction[J]. Laser & Optoelectronics Progress, 2016(8):7-13. ]
[18]	惠振阳,程朋根,官云兰,等.机载LiDAR点云滤波综述[J].激光与光电子学进展,2018(6):7-15.
	[ Hui Z Y, Cheng P G, Guan Y L, et al.Review on airborne LiDAR point cloud filtering[J]. Laser & Optoelectronics Progress, 2018(6):7-15. ]
[19]	Vosselman G.Slope based filtering of laser altimetry data[J]. International Archives of Photogrammetry and Remote Sensing, and Spatial Information Sciences,2000,33(B3):935-942.
[20]	Zhang K, Chen S, Whitman D, et al.A progressive morphological filter for removing nonground measurements form airborne LIDAR data[J]. IEEE Transactions on Geoscience and Remote Sensing, 2003,41(4):872-882.
[21]	苗启广,郭雪,宋建锋,等.基于区域预测的LiDAR点云数据形态学滤波算法[J].激光与光电子学进展,2015,52(1):
	011003.
	[ Miao Q G, Guo X, Song J, et al.LiDAR point cloud data with morphological filter algorithm based on region prediction[J]. Laser & Optoelectronics Progress ,2015,52(1):011003. ]
[22]	Long J, Shelhamer E, Darrell T.Fully convolutional networks for semantic segmentation[J]. IEEE Transactions On Pattern Analysis &Machine Intelligence, 2014,39(4):640-651.
[23]	曾妮红,岳迎春,魏占营,等.车载LiDAR点云滤波的改进不规则三角网加密方法[J].测绘科学,2016,41(9):136-139.
	[ Zeng N H, Yue Y C, Wei Z Y, et al.An improved irregular triangular network encryption method of vehicle-borne LiDAR point clouds[J]. Science of Surveying and Mapping, 2016,41(9):136-139. ]

滤波算法	用时/s	第I类误差点占比/%	第Ⅱ类误差点占比/%
改进的渐进形态学滤波	320	1.07	1.22
2 m分辨率CSF滤波	10	3.11	1.73
1 m分辨率CSF滤波	30	1.32	3.14
0.5 m分辨率CSF滤波	90	0.85	6.98

算法	用时/min	第I类误差点占比/%	第Ⅱ类误差点占比/%
改进的渐进形态学滤波	20	1.98	2.33
1 m分辨率CSF滤波	5	6.63	3.24
0.5 m分辨率CSF滤波	15	3.89	4.98