Research on The Method of Individual Tree Segmentation and Individual Tree Volume Estimation Using Backpack LiDAR

Yang LI; Hongbo XU; Chengxing LING; Xin TIAN; Yanqiu XING; Xin LUO; Zhen GUO; Shuxin CHEN; Haiyi WANG

doi:10.11873/j.issn.1004-0323.2025.3.0568

Remote Sensing Technology and Application, Volume. 40, Issue 3, 568(2025)

Research on The Method of Individual Tree Segmentation and Individual Tree Volume Estimation Using Backpack LiDAR

Yang LI¹, Hongbo XU², Chengxing LING³, Xin TIAN³, Yanqiu XING^1、*, Xin LUO³, Zhen GUO¹, Shuxin CHEN³, and Haiyi WANG³

¹Forest Operations and Environment Research Center， College of Mechanical and Electrical Engineering， Northeast Forestry University， Harbin150040，China

²Xiongan Xiongchuang Digital Technology CO. ，LTD， Xiongan071000，China

³Institute of Forest Resource Information Techniques， Chinese Academy of Forestry， Beijing100091， China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(18)

Fig. 1. Study area and distribution map of sample plots

Download full size

View in Article

Fig. 2. Data acquisition route planning

Download full size

View in Article

Fig. 3. The flowchart of research

Download full size

View in Article

Fig. 4. The examples of the individual tree point cloud from the Backpack LiDAR

Download full size

View in Article

Fig. 5. Schematic diagram of the layered spatial clustering algorithm

Download full size

View in Article

Fig. 6. The procedure of the improved K-means hierarchical clustering algorithm

Download full size

View in Article

Fig. 7. Comparison between the estimate parameters extracted from the point cloud and field survey

Download full size

View in Article

Fig. 8. Estimated individual tree volume of the Scheme 1

Download full size

View in Article

Fig. 9. Pearson correlation coefficient of the model parameters

Download full size

View in Article

Fig. 10. Estimated individual tree volume of Scheme 2（a）

Download full size

View in Article

Fig. 11. Estimated individual tree volume of Scheme 2（b）

Download full size

View in Article

Table 1. Table 1Sample plot information table

View table

View in Article

Table 1. Table 1Sample plot information table

样地号	样地尺寸/m	样地主要树种	树木总数	有效棵树
JD01	20×30	杉木、马尾松、浙江楠	41	41
JD02	25×25	水杉、麻栎、浙江楠	52	48
JD03	20×30	杉木纯林	57	57
JD04	25×25	油桐、无患子	27	23
JD05	20×30	苦槠、无患子	45	40
JD06	20×30	杉木纯林	62	62
JD07	25×25	马尾松纯林	51	50
JD08	25×25	杉木、木荷	66	60

Table 2. Table 2The main parameters of LiBackpack DGC50 scanning system

View table

View in Article

Table 2. Table 2The main parameters of LiBackpack DGC50 scanning system

参数	具体指标
重量（含相机）	8.6 kg
尺寸	960×318×315 mm
扫描频率（单回波）	600 000 pt/s
激光器	Velodyne VLP-16
激光器扫描方式	16线重复扫描
激光器测距	100 m
LiDAR精度	±3 cm
GNSS定位精度	1 cm+1 ppm

Table 3. Table 3Calculation formula of some height parameters

View table

View in Article

Table 3. Table 3Calculation formula of some height parameters

参数名称	计算公式
H_add	$H_{a a d} = \frac{1}{n} \sum_{i = 1}^{n} \| Z_{i} - H_{m e a n} \|$
H_cv	$H_{c v} = (H_{s t d} / H_{m e a n}) \times 100 %$
H_crr	$H_{c r r} = (H_{m e a n} - H_{m i n}) / (H_{m a x} - H_{m i n})$
H_AIHIQR	$H_{A I H I Q R} = H_{A I H 75 %} - H_{A I H 25 %}$
H_IQR	$H_{I Q R} = H_{75 %} - H_{25 %}$
H_sq	$H_{s q} = \sqrt[]{(\sum_{i = 1}^{n} Z_{i}^{2}) / n}$
H_cmc	$H_{c m c} = \sqrt[3]{(\sum_{i = 1}^{n} Z_{i}^{3}) / n}$
Kurtosis	$K u r t o s i s = \frac{\sum_{i = 1}^{n} (Z_{i - H_{m e a n}})^{4}}{(n - 1) {H_{s t d}}^{4}}$
H_sk	$H_{s k} = \frac{\sum_{i = 1}^{n} (Z_{i - H_{m e a n}})^{3}}{(n - 1) {H_{s t d}}^{3}}$

Table 4. Table 4Accuracy evaluation of individual tree segmentation

View table

View in Article

Table 4. Table 4Accuracy evaluation of individual tree segmentation

样地	样地树种	分割方法	TP	FP	FN	P	R	F
JD01	杉木、马尾松、浙江樟	改进的K-means算法	39	0	2	1	0.95	0.97
JD01	杉木、马尾松、浙江樟	分水岭算法	31	3	7	0.91	0.82	0.86
JD02	水杉、麻栎、浙江楠	改进的K-means算法	42	2	8	0.86	0.84	0.85
JD02	水杉、麻栎、浙江楠	分水岭算法	33	6	13	0.85	0.72	0.78
JD03	杉木、马尾松	改进的K-means算法	48	3	6	0.94	0.89	0.91
JD03	杉木、马尾松	分水岭算法	44	4	9	0.92	0.83	0.87
JD04	阔叶混交林	改进的K-means算法	18	3	6	0.86	0.75	0.80
JD04	阔叶混交林	分水岭算法	17	6	4	0.74	0.81	0.77
JD05	阔叶混交林	改进的K-means算法	25	5	15	0.83	0.63	0.72
JD05	阔叶混交林	分水岭算法	26	6	13	0.81	0.67	0.73
JD06	杉木	改进的K-means算法	55	2	5	0.96	0.92	0.94
JD06	杉木	分水岭算法	41	9	12	0.82	0.77	0.79
JD07	马尾松	改进的K-means算法	41	3	7	0.93	0.85	0.89
JD07	马尾松	分水岭算法	33	8	10	0.80	0.77	0.78
JD08	杉木、木荷	改进的K-means算法	58	4	11	0.94	0.84	0.89
JD08	杉木、木荷	分水岭算法	45	8	20	0.85	0.69	0.76

Table 5. Table 5Individual tree structure characteristic
View table
View in Article
Table 5. Table 5Individual tree structure characteristic
参数名称含义
Height 树高
DBH 胸径
CD 冠幅
CA 树冠投影面积
CV 树冠体积
GF 间隙率

Table 6. Table 6Filter the resulting parameters

View table

View in Article

Table 6. Table 6Filter the resulting parameters

参数名称	含义	参数名称	含义
Height	树高	H_cv	变异系数
DBH	胸径	Kurtosis	峰度
CA	树冠投影面积	H_max	最大值
CV	树冠体积	H_99%	高度百分位数四分位数间距
GF	间隙率	H_sq	二次幂平均
H_aad	平均绝对偏差	Variance	方差
AIH_IQ	累积高度百分位数四分位数间距	Density_8（9）	第8层和第9层密度变量

Table 7. Table 7Estimated the sampled plots’ stem volume

View table

View in Article

Table 7. Table 7Estimated the sampled plots’ stem volume

样地号	实测蓄积量 /（m³·hm^-2）	预测蓄积量 /（m³·hm^-2）	误差 /（m³）	精度/%
JD01	111.83	117.17	5.33	95.45
JD02	243.04	222.56	-20.48	90.80
JD03	216.00	205.67	-10.33	94.98
JD04	33.12	36.32	3.20	91.19
JD05	159.33	170.67	11.33	93.36
JD06	349.33	338.50	-10.83	96.80
JD07	148.96	153.92	4.96	96.78
JD08	180.64	170.88	-9.76	94.29

Tools

Get Citation

Copy Citation Text

Yang LI, Hongbo XU, Chengxing LING, Xin TIAN, Yanqiu XING, Xin LUO, Zhen GUO, Shuxin CHEN, Haiyi WANG. Research on The Method of Individual Tree Segmentation and Individual Tree Volume Estimation Using Backpack LiDAR[J]. Remote Sensing Technology and Application, 2025, 40(3): 568

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites