Information Extraction and Security Risk Assessment of Street Trees Based on Vehicle-Borne LiDAR Data

doi:10.13466/j.cnki.lyzygl.2022.05.012

Abstract

Abstract:

Rapid and accurate acquisition of the structural characteristics and safety risk status of urban street trees based on LiDAR point cloud data is of great significance to assist smart city management.In order to solve the problem that LiDAR point cloud data is difficult to segment regions with unclear morphological characteristics in street tree parameter acquisition,a individual tree extraction method based on circular index of trunk center point was proposed.Firstly,it obtained the sliced point cloud of the trunk layer according to the elevation information,then,segmented the sliced data based on the improved DBSCAN clustering algorithm.Secondly,it identified the trunk through the morphological characteristics of the ground features and obtained the central point,so as to complete the extraction through the circular index method based on the central point.Finally,it combined with the risk matrix method to evaluate the safety risk of the stability and traffic impact of the street trees in the study area.The results showed that the proposed individual tree extraction method could effectively improve the segmentation accuracy of individual trees of street trees in areas with unclear morphological characteristics,and accurately obtain the structural parameter information such as the number,shape and position of street trees;The safety risk assessment found that the stability of most street trees and the risk of traffic impact in the study area werein a level I negligible risk state,but there were some street trees whose stability and risk of traffic impact were level II and level III.These trees were mainly distributed in the area with dense and interlaced roadside trees in the study area.The results can provid corresponding decision support for relevant departments to monitor street trees in a timely and effective manner.

Key words: vehicle-borne LiDAR, extraction of information, street trees, security risk assessment

CLC Number:

P225.1
TN249

MU Tianbao, WU Linna, ZHANG Haitao, ZHANG Han. Information Extraction and Security Risk Assessment of Street Trees Based on Vehicle-Borne LiDAR Data[J]. FOREST RESOURCES WANAGEMENT, 2022, (5): 91-98.

Figures/Tables 13

Fig.1

Fig.2

Fig.3

Fig.4

Fig.5

Tab.1

Tab.2

Fig.6

Tab.3

Fig.7

Fig.8

Tab.4

Fig.9

References 29

[1]	Prieur J-F, St-Onge B, Fournier R A, et al. A comparison of three airborne laser scanner types for species identification of individual trees[J]. Sensors, 2022, 22(1):35-57. doi: 10.3390/s22010035
[2]	刘欢, 王玥, 艾明耀. 融合LiDAR点云与高分影像的单木检测方法研究[J]. 地理空间信息, 2021, 19(10):1-4.
[3]	皋厦, 申鑫, 代劲松, 等. 结合LiDAR单木分割和高光谱特征提取的城市森林树种分类[J]. 遥感技术与应用, 2018, 33(6):1073-1083.
[4]	张玥焜, 余文杰, 赵习之, 等. 基于机载激光雷达点云的交互式树木分割与建模方法研究[J]. 图学学报, 2021, 42(4):599-607.
[5]	Gupta A, Byrne J, Moloney D, et al. Tree annotations in LiDAR data using point densities and convolutional neural networks[J]. IEEE Transactions on Geoscience and Remote Sensing, 2019, 58(2):971-981. doi: 10.1109/TGRS.2019.2942201
[6]	C$\check{a}$eanu M, Ciubotaru A. The effect of LiDAR sampling density on DTM accuracy for areas with heavy forest cover[J]. Forests, 2021, 12(3):265-285. doi: 10.3390/f12030265
[7]	赵沛冉, 管海燕, 李迪龙, 等. 利用样本生成方法进行机载多光谱LiDAR数据深度学习分类[J]. 测绘通报, 2021, 12(12):16-21.
[8]	You Hangkai, Li Shihua, Xu Yifan, et al. Tree extraction from airborne laser scanning data in urban areas[J]. Remote Sensing, 2021, 13(17):3428-3446. doi: 10.3390/rs13173428
[9]	谷志新, 裴方睿. 基于多层K-means在森林点云中的单木识别算法[J]. 林业资源管理, 2022(1):124-131.
[10]	Michałowska M, Rapiński J. A review of tree species classification based on airborne LiDAR data and applied classifiers[J]. Remote Sensing, 2021, 13(3):353-379. doi: 10.3390/rs13030353
[11]	Puliti S, Breidenbach J, Astrup R. Estimation of forest growing stock volume with UAV laser scanning data:can it be done without field data?[J]. Remote Sensing, 2020, 12(8):1245-1264. doi: 10.3390/rs12081245
[12]	熊艳, 高仁强, 徐战亚. 机载LiDAR点云数据降维与分类的随机森林方法[J]. 测绘学报, 2018, 47(4):508-518.
[13]	Pvd A, Dha B. Risk assessment concerning urban ecosystem disservices:The example of street trees in Berlin,Germany[J]. Ecosystem Services, 2019, 40(10):31-44.
[14]	宋婷. 城市行道树健康检测评价及应急管理体系构建[D]. 上海: 上海应用技术大学, 2021.
[15]	贾益兴, 雷杰, 黄颂谊. 广州市环市路行道树树木安全风险评价与管理探析[J]. 广东园林, 2021, 43(4):93-96.
[16]	陈文静, 曹雨秋, 张慧会, 等. 法桐行道树生长健康评价——以南京市热河路为例[J]. 中南林业科技大学学报, 2022(8):176-190.
[17]	宋平. 基于VTA-风险矩阵法的上海市行道树安全风险评价研究[D]. 上海: 上海应用技术大学, 2021.
[18]	吴悠绿, 肖鹏峰, 刘豪, 等. 基于因子分析法与k-means聚类的行道树安全风险评价[J/OL]. 西南林业大学学报:自然科学,(2022-05-17)[2022-07-28].http://kns.cnki.net/kcms/detail/53.1218.s.20220517.1015.004.html.
[19]	李永强, 李鹏鹏, 董亚涵, 等. 车载LiDAR点云数据中杆状地物自动提取与分类[J]. 测绘学报, 2020, 49(6):724-735.
[20]	邢旭东, 吕现福, 王旭东, 等. 一种基于分层自适应移动曲面拟合机载LiDAR点云数据滤波方法[J]. 测绘与空间地理信息, 2016, 39(1):128-130.
[21]	张小红. 机载激光雷达测量技术理论与方法[M]. 武汉: 武汉大学出版社, 2007.
[22]	Zhu Xiaoxiao, Wang Cheng, XI Xiaohuan, et al. Hierarchical threshold adaptive for point cloud filter algorithm of moving surface fitting[J]. Acta Geodaetica et CartographicaSinica, 2018, 47(2):153-160.
[23]	Li Shanshan. An improved DBSCAN algorithm based on the neighbor similarity and fast nearest neighbor query[J]. IEEE Access, 2020, 8(3):47468-47476. doi: 10.1109/ACCESS.2020.2972034
[24]	Roberto Ferrara, Salvatore G. P.Virdis, Andrea Ventura, et al. An automated approach for wood-leaf separation from terrestrial LIDAR point clouds using the density based clustering algorithm DBSCAN[J]. Agricultural and Forest Meteorology, 2018, 16(5):434-444.
[25]	Latella M, Sola F, Camporeale C. A density-based algorithm for the detection of individual trees from LiDAR data[J]. Remote Sensing, 2021, 13(2):322-344. doi: 10.3390/rs13020322
[26]	董亚涵, 李永强, 李鹏鹏, 等. 基于车载LiDAR点云的杆状地物分类研究[J]. 测绘工程, 2019, 28(6):58-63.
[27]	Ana PaulaCoelho-Duarte, Gustavo Daniluk-Mosquera, Virginia Gravina, et al. Tree risk assessment:component analysis of six visual methods applied in an urban park,Montevideo,Uruguay[J]. Urban Forestry & Urban Greening, 2021, 59(2):70-79.
[28]	宋婷, 贺坤, 王本耀, 等. 一种行道树健康状况评价方法及系统:中国,CN111639877A[P]. 2020-09-08.
[29]	贺坤, 宋平, 王本耀, 等. 上海城市行道树安全风险评价研究[J]. 中国园林, 2021, 37(9):106-111.

评价类型	评价种类	评价指标	指标定义	评分标准
稳定性风险	可能性评价 (P)	树冠偏冠状况 (P₁)	树冠生长不均匀导致的树冠偏向一边的情况	1(可忽略)	2(轻)	3(中)	4(重)
稳定性风险	可能性评价 (P)	树干倾斜度>10° (P₂)	树干与垂直方向的夹角,夹角越大倾斜程度越大	1(可忽略)	2(轻)	3(中)	4(重)
	严重性评价 (S)	胸径粗细 (S₁)	树干地表胸高处的直径	1(低危)	2(中危)	3(高危)	4(极危)
对交通影响风险	可能性评价 (P)	分支点高度 (P₃)	从地面至树木第一层活分支点的高度	1(可忽略)	2(轻)	3(中)	4(重)
对交通影响风险	可能性评价 (P)	分支点开张角度 (P₄)	分支点以上区域的枝条与垂直方向所构成的夹角	1(可忽略)	2(轻)	3(中)	4(重)
	严重性评价 (S)	行道树位置 (S₂)	行道树所处道路的位置(包括道路等级、是否处于路口、与道路的距离等)	1(低危)	2(中危)	3(高危)	4(极危)

风险等级(R)	分值	解释说明
Ⅳ	12~16	重大风险(该风险因素非常不安全,并会引起较为严重的后果)
Ⅲ	7~9	中等风险(该风险因素不安全,应立即采取措施控制,并非常有可能发生树木倒塌或者其他风险事件)
Ⅱ	4~6	可接受的风险(该风险因素比较安全,担忧潜在风险,应引起注意)
Ⅰ	1~3	可忽略的风险(该风险因素基本安全,但不排除其存在一定的安全隐患,应当继续观察)

项目	行道树/株
人工判读	261
算法识别	263
精细化处理结果	243
误差识别差值	20
行道树提取差值	18
识别率/%	93.10

风险等级	稳定性风险	对交通影响风险
Ⅰ(可忽略的风险)	236	224
Ⅱ(可接受的风险)	6	15
Ⅲ(中等风险)	1	4
Ⅳ(重大风险)	0	0