Extraction of the height of Picea schrenkiana var. tianshanica and inversion of accumulation volume based on airborne lidar images

doi:10.6048/j.issn.1001-4330.2023.04.021

Abstract

Abstract:

【Objective】The data that will be obtained is for building model to invert the accumulation of Picea schrenkiana var. tianshanica in the hope of providing a reference for the ecological restoration and scientific management of Picea schrenkiana var. tianshanica after the implementation of the natural forest protection project.【Methods】To extract the image elevation by using Picea schrenkiana var. tianshanica in the Tiangeer Forest Park on the northern slope of the central Tianshan Mountains in Xinjiang as the research object, and the airborne lidar aerial image and each wooden ruler of the sample plot as the data source. Point cloud classification and Kriging interpolation method were used to extract the height data of the UAV image to obtain the height of the Picea schrenkiana var.tianshanica, and at the same time to establish the DBH tree height model based on the measured data of the sample site, and finally the Picea schrenkiana var. tianshanica was inverted according to the DBH tree height model forest stand volume. 【Results】The UAV image had a high resolution. After point cloud classification, the average accuracy of the tree height extracted by the Kriging interpolation method reached 88.42%. By constructing the DBH tree height model, the power function curve model fit was obtained. The highest, R² was 0.730, combined with the binary volume formula, based on the comparison of the estimated volume of the UAV image with the measured volume of the sample site, the accuracy reached 87.66%.【Conclusion】The Kriging interpolation method has a good effect on the extraction of the height information of the Tianshan spruce forest. The establishment of the model of diameter of DBH makes up for the defect that the UAV cannot directly measure the diameter of the tree, and then inverts the accumulation of the Picea schrenkiana var. tianshanica. This provides a good reference for the restoration of natural forests in Xinjiang after the implementation of the “Natural Forest Protection” project.

Key words: airborne lidar; Picea schrenkiana var. tianshanica; tree height; accumulation

摘要：

【目的】研究提取影像高程数据建立模型反演天山云杉林分蓄积量,获得便捷、快速提取森林蓄积信息的技术方法,为研究山地天然林精准监测与评价提供技术途径。【方法】以新疆天山中部北坡天格尔森林公园天山云杉(Picea Schrenkiana var. tianshanica)为研究对象,机载激光雷达航拍影像与样地每木检尺为数据源,使用点云分类与克里金插值法对激光雷达影像高程数据进行提取获得天山云杉树高,根据样地实测数据构建胸径-树高模型,并根据胸径-树高模型天山云杉林林分蓄积量进行反演。【结果】激光雷达影像分辨率较高,经过点云分类后,采用克里金插值法提取的树高平均精度可达89.64%,幂函数曲线模型拟合度最高,R²为0.908,结合二元材积公式,基于激光雷达影像估测蓄积量与样地实测蓄积量对比,精度达到87.43%。【结论】采用克里金插值法对天山云杉林树高信息的提取效果较好,建立胸径-树高模型弥补了激光雷达不能对胸径直接测量的缺陷,反演天山云杉林林分蓄积量,该模型可满足对新疆山地天然林数字经营管理的标准。

关键词: 机载激光雷达, 天山云杉, 树高, 蓄积量

CLC Number:

S771

QU Yanbin, WANG Zhenxi, HU Tianqi, DONG Wei, CHEN Zhe. Extraction of the height of Picea schrenkiana var. tianshanica and inversion of accumulation volume based on airborne lidar images[J]. Xinjiang Agricultural Sciences, 2023, 60(4): 958-964.

曲延斌, 王振锡, 胡天祺, 董巍, 陈哲. 基于机载激光雷达影像的天山云杉林树高提取及蓄积量反演[J]. 新疆农业科学, 2023, 60(4): 958-964.

Figures/Tables 6

References 18

[1]	Villanueva M, Proupín J, Rodríguez-Añón J A, et al. Energetic characterization of forest biomass by calorimetry and thermal analysis[J]. Journal of Thermal Analysis and Calorimetry, 2011, 104(1): 61-67. DOI URL
[2]	孟宪宇. 2006. 测树学. 北京: 中国林业出版社.
	MENG Xianyu. 2006. Dendrometry. Beijing: China Forestry Publishing House.
[3]	丛沛桐, 祖元刚, 王瑞兰. GIS与ANN整合技术在森林资源蓄积量预测中的应用[J]. 地理科学, 2004, 24(5):591-596.
	CONG Peitong, ZU Yuangang, WANG Ruilan, etc. The application of GIS and ANN integration technology in the prediction of forest reserves[J]. Geographical Sciences, 2004, 24(5): 591-596.
[4]	李明辉, 何风华, 刘云, 等. 天山云杉种群空间格局与动态[J]. 生态学报, 2005, 25(5): 1000-1006.
	LI Minghui, HE Fenghua, LIU Yun, et al. Spatial distribution pattern of tree individuals in the Picea Schrenkiana var tianshanica[J]. Acta Ecologica Sinica, 2005, 25(5): 1000-1006.
[5]	王雅佩, 王振锡, 刘梦婷, 等. 基于无人机影像天山云杉林主伐迹地提取研究[J]. 新疆农业科学, 2019, 56(7): 1315-1327. DOI
	WANG Yapei, WANG Zhenxi, LIU Mengting, et al. Research on estimation of stand density of regenerate Picea Schrenkiana var tianshanica based on UAV Image[J]. Xinjiang Agricultural Sciences, 2019, 56(7): 1315-1327. DOI
[6]	吴娇娇, 张亚红, 杨凯博, 等. 机载激光雷达在林业中的应用[J]. 安徽农业科学. 2016, 44(35): 209-212.
	WU Jiaojiao, ZHANG Yahong, YANG Kaibo, et al. Application of airborne lidar in forestry[J]. Journal of Anhui Agricultural Sciences. 2016, 44(35): 209-212.
[7]	Nsset E, Gobakken T. Estimation of above-and below-ground bio-mass across regions of the boreal forest zone using airborne laser[J]. Remote Sensing of Environment, 2008, 112(60): 3079-3090. DOI URL
[8]	洪奕丰, 张守功, 陈伟, 等. 基于机载激光雷达的落叶松组分生物量反演[J]. 林业科学研究, 2019, 32(5): 83-90.
	HONG Yifeng, ZHANG Shougong, CHEN Wei, et al. Retrieval of larch component biomass based on airborne lidar[J]. Forestry Science Research, 2019, 32(5): 83-90.
[9]	Naesset E. Predicting forest stand characteristics with airborne scanning lase using a practical two-stage procedure and field data[J]. Remote Sensing of Environment, 2002, 80(1): 88-99. DOI URL
[10]	穆喜云, 张秋良, 刘清旺, 等. 基于机载LiDAR数据的林分平均高及郁闭度反演[J]. 东北林业大学学报, 2015, 43(9): 84-89.
	MU Xiyun, ZHANG Qiuliang, LIU Qingwang, et al. Inversion of forest height and canopy closure using airborne lidar data Inversion of average stand height and canopy closure based on airborne LiDAR data[J]. Journal of Northeast Forestry University, 2015, 43(9): 84-89.
[11]	Wallace L, Lucieer A, Watson C, et al. Development of a UAV-Li-DAR system with application to forest inventory[J]. Remote Sensing, 2012, 4: 1519-1543. DOI URL
[12]	Argialis A, Erfanifard Y, Dehghani M, et al. Woodland mapping atsingle-tree levels using object-oriented classification of Unmanned Aerial Vehicle images[J]. Remote Sensing and Spatial Information Sciences, 2017,(4): 43-49.
[13]	吕金城, 王振锡, 杨勇强. 基于无人机影像的天山云杉林树高提取及蓄积量的反演[J]. 新疆农业科学, 2021, (10):2-9.
	LU Jincheng, WANG Zhenxi, YANG Yongqiang, etc. The height extraction and stock volume inversion of Tianshan spruce forest based on UAV images[J]. Xinjiang Agricultural Sciences, 2021, (10): 2-9.
[14]	KRAUS K, PFEIFER N. Determination of terrain models in wooded areas with airborne laser scanner data[J]. ISPRS Journal of Photogrammetry & Remote Sensing, 1998, 53(4):193-203.
[15]	庞勇, 李增元, 陈尔学, 等. 激光雷达技术及其在林业上的应用[J]. 林业科学, 2005, 41(3):129-136.
	PANG Yong, LI Zengyuan, CHEN Erxue, et al. Lidar technology and its application in forestry[J]. Forestry Science, 2005, 41(3):129-136.
[16]	马利群, 李爱农. 激光雷达在森林垂直结构参数估算中的应用[J]. 世界林业研究, 2011, 24(1):41-45.
	MA Liqun, LI Ainong. The application of lidar in the estimation of forest vertical structure parameters[J]. World Forestry Research, 2011, 24(1):41-45.
[17]	何祺胜, 陈尔学, 曹春香, 等. 基于LIDAR数据的森林参数反演方法研究[J]. 地球科学进展, 2009(7):748-755.
	He Qisheng, Chen Erxue, Cao Chunxiang, et al. Research on forest parameter inversion method based on LIDAR data[J]. Progress in Earth Science, 2009 (7): 748-755
[18]	马骧, 姜文龙, 闫保银, 等. 基于机载激光雷达的国外松片林蓄积量反演研究[J]. 江苏林业科技, 2021, 48( 3) : 32-37.
	MA Xiang, JIANG Wenlong, YAN Baoyin, et al. Foreign Pine Forest Volume Inversion Research Based on Airborne Lidar[J]. Jiangsu Forestry Science and Technology, 2021, 48(3): 32-37.

模型 Model	平方和 Sum of squares	df	均方 Mean square	F	Sig.	R²
回归 Return	2 334.187	1	2 334.187	1 150.511	0.000	0.886
残差 Residual	300.266	148	2.029

模型 Model	平方和 Sum of squares	df	均方 Mean square	F	Sig.	R²
回归 Return	2 334.187	1	2 334.187	1 150.511	0.000	0.886
残差 Residual	300.266	148	2.029

方程Model	R²	F	df1	df2
线性 Linear function(D=1.191×H-0.437)	0.901	891.645	1	98
对数 Logarithmic function(D=-40.431+21.547log(H))	0.878	702.104	1	98
倒数 Reciprocal function(D=41.808-341.66/H)	0.79	368.721	1	98
二次 Quadratic function(D=-1.969+1.358×H-0.004×H²)	0.902	446.552	2	97
三次 Cubic function(D=-1.913+1.348×H-0.004×H²-0.0009×H³)	0.905	306.307	3	96
复合 Function of function(D=7.337×1.058^H)	0.884	743.861	1	98
幂 Power function(D=(H×1.282)^0.872)	0.908	972.173	1	98
S曲线 S-curve(D=exp(4.030-17.037/H))	0.867	640.363	1	98
增长 Increasing function(D=exp(1.993+0.056×H))	0.884	743.861	1	98
指数 Exponential function(D=7.337×exp(0.056×H))	0.884	743.861	1	98
Logistic (D=1/(0.136×0.946^H))	0.884	743.861	1	98

方程Model	R²	F	df1	df2
线性 Linear function(D=1.191×H-0.437)	0.901	891.645	1	98
对数 Logarithmic function(D=-40.431+21.547log(H))	0.878	702.104	1	98
倒数 Reciprocal function(D=41.808-341.66/H)	0.79	368.721	1	98
二次 Quadratic function(D=-1.969+1.358×H-0.004×H²)	0.902	446.552	2	97
三次 Cubic function(D=-1.913+1.348×H-0.004×H²-0.0009×H³)	0.905	306.307	3	96
复合 Function of function(D=7.337×1.058^H)	0.884	743.861	1	98
幂 Power function(D=(H×1.282)^0.872)	0.908	972.173	1	98
S曲线 S-curve(D=exp(4.030-17.037/H))	0.867	640.363	1	98
增长 Increasing function(D=exp(1.993+0.056×H))	0.884	743.861	1	98
指数 Exponential function(D=7.337×exp(0.056×H))	0.884	743.861	1	98
Logistic (D=1/(0.136×0.946^H))	0.884	743.861	1	98