基于增强Frost局部滤波及单木距离图重构标记的CHM树冠分割

doi:10.12302/j.issn.1000-2006.202209011

国家林草科技领军期刊
中国精品科技期刊
中国高校百佳科技期刊
江苏省新闻出版政府奖期刊奖
RCCSE林学权威期刊（A+）
CSCD核心期刊
Scopus数据库收录期刊
中文核心期刊
SCD核心期刊

作者加群：102861116

微信公众号：南京林业大学学报

高级检索

南京林业大学学报（自然科学版） ›› 2023, Vol. 47 ›› Issue (5): 9-18.doi: 10.12302/j.issn.1000-2006.202209011

所属专题：林草计算机应用研究专题

• 专题报道：林草计算机应用研究专题（执行主编李凤日） • 上一篇下一篇

基于增强Frost局部滤波及单木距离图重构标记的CHM树冠分割

张华聪¹^,²^,³(), 谭新建³(), 喻龙华³, 厉月桥³, 陈永富¹^,², 刘仁³, 张怀清¹^,²^,^*()

1.中国林业科学研究院资源信息研究所,北京 100091
2.国家林业和草原局林业遥感与信息技术重点实验室,北京 100091
3.中国林业科学研究院亚热带林业实验中心,江西新余 336600

收稿日期:2022-09-05 修回日期:2022-10-25 出版日期:2023-09-30 发布日期:2023-10-10
作者简介:张华聪(718736170@qq.com),博士生,负责论文初稿撰写;谭新建(bjtan@caf.ac.cn),教授级高级工程师,负责论文修改。
基金资助:
中国林业科学研究院资源信息研究所基本科研业务费专项(CAFYBB2019SZ004);中国林业科学研究院资源信息研究所基本科研业务费专项(CAFYBB2021ZE005)

Crown segmentation of CHM based on the enhanced frost local filtering and distance map reconstruction

ZHANG Huacong¹^,²^,³(), TAN Xinjian³(), YU Longhua³, LI Yueqiao³, CHEN Yongfu¹^,², LIU Ren³, ZHANG Huaiqing¹^,²^,^*()

1. Institute of Forest Resource Information Techniques, CAF, Beijing 100091, China
2. Key Laboratory of Remote Sensing and Information, NFGA, Beijing 100091, China
3. Experimental Enter of Subtropical Forestry, CAF, Xinyu 336600, China

Received:2022-09-05 Revised:2022-10-25 Online:2023-09-30 Published:2023-10-10

摘要/Abstract

摘要：

【目的】利用增强Frost局部滤波和单木距离图重构标记技术对冠层高度模型(CHM)进行分割,以提高无人机激光雷达在单木树冠分割的精度和效率。【方法】选取江西分宜山下实验林场阔叶混交林、针叶混交林和针阔混交林3个不同类型的林分为研究对象,以无人机激光雷达数据为数据源,构建CHM。针对高分辨率CHM树冠区域孔隙较多的问题,利用增强Frost局部滤波处理优化CHM,优化结果与不同滤波方法进行了比较分析;然后应用距离图重构标记分割技术对增强Frost局部滤波优化后的0.1、0.2、0.5及1.0 m分辨率的CHM进行分割与分析;最后确定最佳分辨率的CHM,并其将分割结果与同等分辨率下分水岭算法以及点云分割均值偏移算法结果进行比较。【结果】采用增强Frost局部滤波处理的CHM优化效果显著,在有效抑制树冠噪音的同时,也能较好地保留图像细节信息。0.2 m分辨率的CHM分割效果最佳。距离图重构标记分割方法分割针叶混交林、针阔混交林、阔叶混交林3种不同林分类型的分割精度(OA)分别为0.96、0.84、0.75;根据树冠分割结果计算单木冠幅,冠幅估测的决定系数R²分别为0.83、0.82、0.71。【结论】基于增强Frost局部滤波及距离图重构标记技术可实现对激光点云CHM的单木分割和树冠估算,能够满足森林调查和监测的基本需求。

关键词: 无人机激光点云, 林冠高度模型(CHM), 增强Frost滤波, 距离图重构标记, 树冠分割

Abstract:

【Objective】We used the enhanced Frost local filtering and single-tree distance map reconstruction marking technology to segment a canopy height model (CHM), improving the accuracy and efficiency of unpiloted aerial system (UAV)-light detection and ranging (LiDAR) segmentation in single-tree crowns.【Method】 We selected three forest types - coniferous mixed, coniferous-broad mixed, and broad-leaved mixed - in the Shanxia Experimental Forest Farm of Fenyi, Jiangxi Province. We then used UAV-LiDAR data to construct the CHM. To combat the increased pores in the crown area of the high-resolution CHM, we used the enhanced Frost local filtering to optimize the CHM and results were compared with different filtering methods. Next we applied the distance map reconstruction marker segmentation technology to segment and analyze the CHM-with resolutions of 0.1, 0.2, 0.5 and 1.0 m after optimization of the enhanced Froest local filter. Finally, we determined the CHM with the optimal resolution, and compared segmentation results with that of a watershed algorithm with the same resolution and mean-shift segmentation algorithm. 【Result】 Applying the enhanced Frost local filter indeed optimized the CHM-preserving image details while suppressing phase crown noise. A resolution of 0.2 m performed best for the CHM segmentation. An overall accuracy of 0.96, 0.84 and 0.75 was observed for coniferous mixed, coniferous-broad mixed, and broad-leaved mixed forests, respectively. The crown width of a single tree was calculated according to crown segmentation results, and the R² estimated at 0.83, 0.82 and 0.71, respectively. 【Conclusion】Through the enhanced Frost local filtering and distance map reconstruction marking technology, the single-tree segmentation and crown estimation of laser point cloud CHMs can be realized, meeting key requirements of forest surveys and monitoring.

Key words: UAV laser point cloud, cannopy height model(CHM), enhance Frost, distance map marker and reconstruction, crown segmentation

中图分类号:

S758.5

张华聪,谭新建,喻龙华,等. 基于增强Frost局部滤波及单木距离图重构标记的CHM树冠分割[J]. 南京林业大学学报（自然科学版）, 2023, 47(5): 9-18.

ZHANG Huacong, TAN Xinjian, YU Longhua, LI Yueqiao, CHEN Yongfu, LIU Ren, ZHANG Huaiqing. Crown segmentation of CHM based on the enhanced frost local filtering and distance map reconstruction[J].Journal of Nanjing Forestry University (Natural Science Edition), 2023, 47(5): 9-18.DOI: 10.12302/j.issn.1000-2006.202209011.

图/表 7

图1

图2

表1

图3

图4

表2

图5

参考文献 34

[1]	解宇阳, 王彬, 姚扬, 等. 基于无人机激光雷达遥感的亚热带常绿阔叶林群落垂直结构分析[J]. 生态学报, 2020, 40(3):196-207.
	XIE Y Y, WANG B, YAO Y, et al. Quantification of vertical community structure of subtropical evergreen broadleaved forest community using UAV-Lidar data[J]. Acta Ecologica Sinica, 2020, 40(3):196-207. DOI:10.5846/stxb201809282112.
[2]	陈宗铸, 杨琦, 雷金睿, 等. 基于激光雷达数据的热带森林冠高模型生成及平均树高估计[J]. 中南林业科技大学学报, 2018, 38(7):1-7.
	CHEN Z T, YANG Q, LEI J R, et al. Generation of tropical forest canopy height model and estimation of average tree height based on lidar data[J]. Journal of Central South University of Forestry and Technology, 2018, 38(7):1-7. DOI:10.14067/j.cnki.1673-923x.2018.07.001.
[3]	巨一琳, 姬永杰, 黄继茂, 等. 联合LiDAR和多光谱数据森林地上生物量反演研究[J]. 南京林业大学学报(自然科学版), 2022, 46(1):58-68.
	JU Y L, JI Y J, HUANG J M, et al. Inversion of forest aboveground biomass using combination of LiDAR and multispectral data[J]. Journal of Nanjing Forestry University(Natural Sciences Edition), 2022, 46(1):58-68. Doi: 10.12302/j.issn.1000-2006.202109029.
[4]	ALVITES C I, SANTOPUOLI G, MAESANO M, et al. Unsupervised algorithms to detect single trees in a mixed-species and multi-layered Mediterranean forest using LiDAR data[J]. Canadian Journal of Forest Research, 2021, 51(12):1766-1780. DOI:10.1139/cjfr-2020-0510.
[5]	JIANG X, LI G, LU D, et al. Modelingforest aboveground carbon density in the brazilian amazon with integration of MODIS and airborne LiDAR data[J]. Remote Sensing, 2020, 12(20): 3330. DOI:10.3390/rs12203330.
[6]	张冬, 云挺, 薛联凤, 等. 基于力场的点云树木骨架提取方法[J]. 南京林业大学学报(自然科学版), 2016, 40(2):160-166.
	ZHANG D, YUN T, XUE L F, et al. Skeleton extraction for point cloud trees based on force field[J]. Journal of Nanjing Forestry University(Natural Sciences Edition), 2016, 40(2):160-166. DOI:10.3969/j.issn.1000-2006.2016.02.027.
[7]	BLACKBURN R C, BUSCAGLIA R, MEADOR A. Mixtures of airborne lidar-based approaches improve predictions of forest structure[J]. Canadian Journal of Forest Research, 2021, 51(8):1106-1116. DOI:10.1139/cjfr-2020-0506
[8]	李文娟, 赵传燕, 别强, 等. 基于机载激光雷达数据的森林结构参数反演[J]. 遥感技术与应用, 2015, 30(5):917-924.
	LI W J, ZHAO C Y, BIE Q, et al. Retrieval of the forest structural parameters using airborne LiDAR data[J]. Remote Sensing Technology and Application, 2015, 30(5):917-924. DOI:10.11873/j.issn.1004-0323.2015.5.0917.
[9]	VALERO S, KNAPP D E, TOCHON G, et al. On the use of binary partition trees for the tree crown segmentation of tropical rainforest hyperspectral images[J]. Remote Sensing of Environment, 2015, 159:318-331.DOI:10.1016/j.rse.2014.12.020.
[10]	王辉, 韩娜娜, 吕程序, 等. 基于Mask R-CNN的单株柑橘树冠识别与分割[J]. 农业机械学报, 2021, 52(5):169-174.
	WANG H, HAN N N, LYU C X, et al. Recognition and segmentation of individual citrus tree crown based on Mask R-CNN[J]. Transactions of the Chinese Society of Agricultural Machinery, 2021, 52(5):169-174. DOI:10.6041/j.issn.1000-1298.2021.05.018.
[11]	程晓菲, 武刚. 基于高分辨率卫星影像的CV模型单木定位法[J]. 南京林业大学学报(自然科学版), 2022, 46(5):143-151.
	CHEN X F, WU G. Locating individual tree from high resolution satellite images based on CV model[J]. Journal of Nanjing Forestry University(Natural Sciences Edition), 2022, 46(5):143-151. DOI:10.12302/j.issn.1000-2006.202102020.
[12]	孙振峰, 张晓丽, 李霓雯. 机载与星载高分遥感影像单木树冠分割方法和适宜性对比[J]. 北京林业大学学报, 2019, 41(11):66-75.
	SUN Z F, ZHANG X L, LI N W. Comparison of individual tree crown extraction method and suitability of airborne and spaceborne high-resolution remote sensing images[J]. Journal of Beijing Forestry University, 2019, 41(11):66-75. DOI:10.13332/j.1000-1522.20180446.
[13]	YANG L, SHENG Y H, WANG B. 3D reconstruction of building facade with fused data of terrestrial LiDAR data and optical image[J]. Optik, 2016, 127(4):2165-2168. DOI:10.1016/j.ijleo.2015.11.147.
[14]	MAN Q, DONG P, GUO H. Pixel and feature-level fusion of hyperspectral and lidar data for urban land-use classification[J]. International Journal of Remote Sensing, 2015, 36(5/6):1618-1644. DOI:10.1080/01431161.2015.1015657.
[15]	LI Y, JAVIER I G. LiDAR for autonomous driving: the principles, challenges, and trends for automotive lidar and perception systems[J]. IEEE Signal Processing Magazine, 2020, 37(4):50-61. DOI:10.1109/MSP.2020.2973615.
[16]	FABAIN D, REIK L, FELIX M, et al. Simulating imaging spectrometer data: 3D forest modeling based on LiDAR and in situ data[J]. Remote Sensing of Environment, 2014, 152:235-250. DOI: 10.1016/j.rse.2014.06.015.
[17]	徐志扬, 刘浩栋, 陈永富, 等. 基于无人机LiDAR的杉木树冠上部外轮廓模拟与可视化研究[J]. 林业科学研究, 2021, 34(4):9.
	XU Z Y, LIU H D, CHEN Y F, et al. Outer upper crown profile aimulation and visualization for Cunninghamia lanceolata based on UAV-borne LiDAR data[J]. Forest Research, 2021, 34(4):9. DOI:10.13275/j.cnki.lykxyj.2021.04.005.
[18]	HUO L, LINDBERG E, HOLMGREN J. Towards low vegetation identification:a new method for tree crown segmentation from LiDAR data based on a symmetrical structure detection algorithm (SSD)[J]. Remote Sensing of Environment, 2022, 270: 112857. DOI:10.1016/j.rse.2021.112857.
[19]	AYREY E, FARVER S, KERSHAW, et al. Layer Stacking: a novel algorithm for individual forest tree segmentation from LiDAR point clouds[J]. Canadian Journal of Remote Sensing, 2017, 43(1):16-27. DOI:10.1080/07038992.2017.1252907.
[20]	HUI Z, LI N, XIA Y, et al. Individual tree extraction from UAV LIDAR point clouds based on self-adaptive mean shift segmentation[J]. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2021(1):25-30. DOI:10.5194/isprs-annals-V-1-2021-25-2021.
[21]	CHEN X, JIANG K, ZHU Y, et al. Individual tree crown segmentation directly from UAV-borne LiDAR data using thePointNet of deep learning[J]. Forests, 2021, 12(2):131. DOI:10.3390/f12020131.
[22]	WANG J, CHEN X, CAO L, et al. Individual rubber tree segmentation based on ground-based LiDAR data and faster R-CNN of deep learning[J]. Forests, 2019, 10(9):793-813. DOI:10.3390/f10090793.
[23]	黄昕晰, 夏凯, 冯海林, 等. 基于无人机影像与Mask R-CNN的单木树冠检测与分割[J]. 林业工程学报, 2021, 6(2):133-140.
	HUANG X X, XIA K, FENG H L, et al. Research on individual tree crown detection and segmentation using UAV imaging and Mask R-CNN[J]. J Fore Eng, 2021, 6(2):133-140.DOI:10.13360/j.issn.2096-1359.202009004.
[24]	VINCENT L, SOILLE P. Watersheds in digital spaces: an efficient algorithm based on immersion simulations[J]. IEEE Computer Society, 1991, 13(6):583-598. DOI:10.1109/34.87344.
[25]	CHEN Q, DENNIS B, GONG P, et al. Isolating individual trees in a savanna woodland using small footprint LIDAR data[J]. Photogrammetric Engineering and Remote Sensing, 2006, 72(8):923-932. DOI:10.14358/PERS.72.8.923.
[26]	王鑫运, 黄杨, 邢艳秋, 等. 基于无人机高密度LiDAR点云的人工针叶林单木分割算法[J]. 中南林业科技大学学报, 2022, 42(8):66-77.
	WANG X Y, HUANG Y, LI D J, et al. The single tree segmentation of UAV high-density LiDAR point cloud data based on coniferous plantations[J]. Journal of Central South University of Forestry and Technology, 2022, 42(8):66-77. DOI:10.14067/j.cnki.1673-923x.2022.08.007.
[27]	李岩, 史泽林, 程坤, 等. 运用激光雷达数据的单木树冠提取算法对帽儿山林场单木参数估测的影响[J]. 东北林业大学学报, 2019, 47(11):61-67.
	LI Y, SHI Z L, CHEN K, et al. Effect of three individual tree crown delineation algorithms on estimating individual tree parameters with LiDAR data[J]. Journal of Northeast Forestry University, 2019, 47(11):61-67. DOI:10.13759/j.cnki.dlxb.2019.11.012.
[28]	张海清, 李向新, 王成, 等. 结合DSM的机载LiDAR单木树高提取研究[J]. 地球信息科学学报, 2021, 23(10):1873-1881. doi: 10.12082/dqxxkx.2021.210030
	ZHANG H Q, LI X X, WANG C, et al. Individual tree height extraction from airborne LiDAR data by combining with DSM[J]. Journal of Geo-information Science, 2021, 23(10):1873-1881. DOI:10.12082/dqxxkx.2021.210030.
[29]	LOPES A, NEZRY E, TOUZI R, et al. Structure detection and statistical adaptive speckle filtering in SAR images[J]. International Journal of Remote Sensing, 1993, 14(9):1735-1758. doi: 10.1080/01431169308953999
[30]	贾惠珍, 王同罕. 基于自适应微调因子的改进 frost 滤波[J]. 计算机工程与设计, 2011, 32(11): 3793-3795.
	JIA H Z, WANG T H. Modified frost filtering based on adaptive parameter[J]. Computer Engineering and Design, 2011, 32(11):3793-3795.
[31]	VINCENT L. Morphological grayscale reconstruction in image analysis: applications and efficient algorithms[J]. IEEE transactions on image processing, 1993, 2(2):176-201. DOI:10.1109/83.217222. pmid: 18296207
[32]	李增元, 庞勇, 刘清旺, 等. 激光雷达森林参数反演技术与方法[M]. 北京: 科学出版社, 2015.
	LI Z Y, PANG Y, LIU Q W, et al. Lidar forest parameter inversion technology and method[M]. Beijing: Science Press, 2015.
[33]	王娟, 张超, 陈巧, 等. 结合无人机可见光和激光雷达数据的杉木树冠信息提取[J]. 西南林业大学学报, 2022, 42(1):133-141.
	WANG J, ZHANG C, CHEN Q, et al. The method of extracting information of Cunninghamia lanceolata crown combined with RGB and LiDAR based on UAV[J]. Journal of Southwest Forestry University. 2022, 42(1):133-141. DOI: 10.11929/j.swfu.202101013.
[34]	汪垚, 张志玉, 倪文俭, 等. 基于机载LiDAR数据的林下地形提取算法比较与组合分析[J]. 北京林业大学学报, 2017, 39(12):25-35.
	WANG Y, ZHANG Z Y, NI W J, et al. Comparison of filter algorithms and combination analysis for DEM extracting based on airborne laser scanning point clouds[J]. Journal of Beijing Forestry University, 2017, 39(12):25-35. DOI: 10.13332/j.1000-1522.20170300.

指标 index	冠层类型 canopy layer type	原始图像 original image	增强Frost enhanced frost	增强Lee enhanced Lee	局部Sigma local sigma	中值滤波 median filtering	本研究 this research
平滑指数(SI) smoothness index	A阔叶混交林	2.38	2.56	2.83	2.36	2.53	2.63
	B针叶混交林	3.06	3.39	3.39	3.13	3.34	3.31
	C针阔混交	2.48	2.66	2.65	2.52	2.58	2.60
	均值	2.64	2.87	2.96	2.67	2.82	2.85
信噪比(SNR) signal to noise ratio	A阔叶混交林	14.33	14.83	14.71	14.33	14.61	14.53
	B针叶混交林	13.25	14.05	14.05	13.35	13.79	13.61
	C针阔混交	13.66	14.11	14.08	13.73	14.17	13.91
	均值	13.75	14.33	14.28	13.77	14.19	14.02
等效视数(ENL) equivalent number of looks	A阔叶混交林	5.70	6.59	8.06	5.63	6.47	6.98
	B针叶混交林	9.46	11.55	11.55	9.90	11.24	11.04
	C针阔混交	6.22	7.14	7.09	6.38	6.72	6.81
	均值	7.13	8.43	8.90	7.31	8.14	8.28
边缘保持指数(EPI) edge protect index	A阔叶混交林		0.44	0.51	0.91	0.45	0.89
	B针叶混交林		0.47	0.48	0.86	0.43	0.85
	C针阔混交		0.48	0.49	0.89	0.44	0.85
	均值		0.46	0.49	0.89	0.44	0.86

冠层类型 canopy layer type	样地序号 sample No.	实测株树 quantity of tree	总体精度(OA) overall accuracy			误判误差 (CE) commission error			漏判误差(OE) omission error
冠层类型 canopy layer type	样地序号 sample No.	实测株树 quantity of tree	本研究 this paper	分水岭 watershed	均值漂移 mean shift	本研究 this paper	分水岭 watershed	均值漂移 mean shift	本研究 this paper	分水岭 watershed	均值漂移 mean shift
A阔叶混交林 broad-leaved mixed forest	1	41	0.71	0.37	0.37	0.27	0.61	0.61	0.02	0.02	0.02
	2	52	0.79	0.42	0.37	0.21	0.54	0.58	0.00	0.04	0.05
	3	41	0.76	0.44	0.32	0.22	0.56	0.66	0.02	0.00	0.02
	均值mean	45	0.75	0.41	0.35	0.23	0.57	0.62	0.02	0.02	0.03
B针叶混交林 coniferous mixed forest	4	67	0.97	0.78	0.66	0.03	0.21	0.31	0.00	0.01	0.03
	5	62	0.94	0.81	0.74	0.05	0.18	0.23	0.02	0.02	0.03
	6	56	0.98	0.84	0.71	0.02	0.16	0.24	0.00	0.00	0.05
	均值mean	62	0.96	0.81	0.70	0.03	0.18	0.26	0.01	0.01	0.04
C针阔混交林 coniferous broad- leaved mixed forest	7	59	0.86	0.69	0.54	0.10	0.29	0.41	0.03	0.02	0.05
	8	65	0.82	0.72	0.58	0.14	0.25	0.38	0.05	0.03	0.04
	9	57	0.84	0.68	0.56	0.11	0.30	0.40	0.05	0.02	0.04
	均值mean	60	0.84	0.70	0.57	0.12	0.28	0.40	0.04	0.02	0.03
总体均值total mean		56	0.85	0.64	0.55	0.13	0.34	0.43	0.02	0.02	0.02

基于增强Frost局部滤波及单木距离图重构标记的CHM树冠分割

Crown segmentation of CHM based on the enhanced frost local filtering and distance map reconstruction

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 7

参考文献 34

相关文章 9

编辑推荐

Metrics

本文评价

作者

读者

审稿

[1]	辛士冬, 姜立春, 穆林. 黑龙江省红松人工林林分乔木层可加性碳储量模型[J]. 南京林业大学学报（自然科学版）, 2022, 46(1): 115-121.
[2]	崔泽宇, 张怀清, 左袁青, 杨廷栋, 刘洋, 张京, 王林龙. 杉木三维模型各方向枝下高分布研究[J]. 南京林业大学学报（自然科学版）, 2022, 46(1): 81-87.
[3]	单凯丽, 臧颢, 潘萍, 宁金魁, 欧阳勋志. 赣中天然闽楠单木冠幅预测模型的研究[J]. 南京林业大学学报（自然科学版）, 2022, 46(1): 88-96.
[4]	王冬至, 胡雪娇, 李大勇, 高雨珊, 李天宇. 基于非线性混合效应模型的针阔混交林地位指数研究[J]. 南京林业大学学报（自然科学版）, 2020, 44(4): 159-166.
[5]	潘磊, 孙玉军, 王轶夫, 陈丽萍, 曹元帅. 基于Sentinel⁃1和Sentinel⁃2数据的杉木林地上生物量估算[J]. 南京林业大学学报（自然科学版）, 2020, 44(3): 149-156.
[6]	周泽宇, 杨绕华, 张玉珍, 黄选瑞, 张志东, 王冬至, 李大勇. 华北落叶松人工林直径分布预测模型构建[J]. 南京林业大学学报（自然科学版）, 2020, 44(2): 117-124.
[7]	李光友,徐建民,王伟,吴世军,朱映安, 王英生,潘建,郭洪英,施庭有. 杂交桉家系在冷凉区优势评价与遗传分析[J]. 南京林业大学学报（自然科学版）, 2017, 41(04): 55-63.
[8]	张密芳,胡曼,李明阳. 基于PALSAR全极化数据的城市森林蓄积量估测[J]. 南京林业大学学报（自然科学版）, 2016, 40(06): 56-62.
[9]	杨铭,王月婷,张晓丽. 基于双源遥感数据的杉木林分蓄积量估测模型研究[J]. 南京林业大学学报（自然科学版）, 2016, 40(05): 107-114.