机载LIDAR数据的树高识别算法与应用分析
作者简介:王轶夫(1990-),男,安徽安庆人,博士生,研究方向为资源环境与生态系统模拟。E-mail:wangyf@lreis.ac.cn
收稿日期: 2014-02-20
要求修回日期: 2014-03-20
网络出版日期: 2014-11-01
基金资助
国家自然科学基金重点项目(91325204)
国家高技术研究发展计划项目(2013AA122003)
科技基础性工作专项(2013FY111600-4)
Study of Factors Impacting the Tree Height Extraction Based on Airborne LIDAR Data
Received date: 2014-02-20
Request revised date: 2014-03-20
Online published: 2014-11-01
Copyright
利用机载激光雷达数据提取天然次生林的树高,旨在探索影响树高提取精度的主要因素。首先,采用高精度曲面建模平差算法(Adjustment Computation of High-accuracy Surface Modeling,HASM-AD)生成研究区不同空间分辨率的数字高程模型(Digital Elevation Model,DEM)、数字地表模型(Digital Surface Model,DSM)和冠层高度模型(Canopy Height Model,CHM);其次,用树顶点识别算法提取林木树高,设置不同树高识别范围,对比分析不同CHM分辨率和不同树高识别范围对树高提取精度的影响;最后,以天涝池流域30个实测样地数据为样本,对提取精度进行检验。结果显示:提取的样地平均树高与实测值具有明显线性相关关系,线性回归系数为0.694;树高识别范围是影响树高提取精度的重要因素,CHM分辨率对其影响较小。研究表明,采用高采样密度的雷达点云数据、正确选择CHM生成方法和改进树顶点识别算法是提高天然次生林树高提取精度的有效途径。
王轶夫 , 岳天祥 , 赵明伟 , 杜正平 , 刘向锋 , 刘爽 , 宋二非 , 孙文正 , 张彦丽 . 机载LIDAR数据的树高识别算法与应用分析[J]. 地球信息科学学报, 2014 , 16(6) : 958 -964 . DOI: 10.3724/SP.J.1047.2014.00958
The purpose of this study is to evaluate the accuracy of extracting average height of natural secondary forest using airborne LIDAR data and to explore the problems that accompany. The DSMs and DEMs with differentspatial resolutions were simulated, by applying HASM-AD algorithm. DSM minus DEM gives CHM, and the tree heights were extracted from CHM. We applied tree vertex recognition algorithm with different recognition scopes. Using 30 measured plot data for verification, we tried to express how CHM spatial revolutionand recognition scope could affect tree height extraction accuracy. Firstly, we produced the 0.5 m resolution of CHM and gave 3 trials with setting the recognition scope radius as 0.5 m, 1.0 m and 1.5 m consecutively. The contrast between the results showed that the number of tree vertices extracted was the largest when the recognition scope radius was set as 0.5 m. The algorithm's ability to recognize tree vertex decreases as recognition scope radius increases. Then, we set the recognition scope radius as 0.5 m unchanged and gave 3 trials in which we extracted tree vertex from different CHM with 3 different resolutions (0.1 m, 0.25 m, 0.5 m). The results showed that the number of tree vertices extracted in 3 trials were close. In other words, the recognition scope radius could hardly influence tree vertex extraction. Finally, we compared the average value of the extracted tree heights in each plot to the average of the measured values. The result showed that they were highly correlated with each other, and the regression coefficient between them was 0.694. In conclusion, the recognition scope radius has great influence on tree vertex extraction, while resolution of CHM has little influence on tree vertex extraction. Increasing the sampling density of LIDAR data, choosing an appropriate CHM simulation method and improving the tree vertex recognition algorithm can increase the accuracy of tree height extraction.
Key words: airborne LIDAR; HASM; tree height; Tianlaochi; natural secondary forest
Tab. 1 Description of plots表1 样地基本信息表 |
样地号 | 森林类型 | 株数 | 平均树高(m) | 最大树高(m) | 最小树高(m) | 平均胸径(cm) | 样地号 | 森林类型 | 株数 | 平均树高(m) | 最大树高(m) | 最小树高(m) | 平均胸径(cm) |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | 祁连圆柏 | 30 | 7.3 | 11.3 | 2.2 | 29.7 | 16 | 青海云杉 | 34 | 11.0 | 21.2 | 2.6 | 15.6 |
2 | 祁连圆柏 | 13 | 8.8 | 12.7 | 2.5 | 33.5 | 17 | 青海云杉 | 27 | 13.0 | 21.6 | 2.4 | 19.8 |
3 | 青海云杉 | 60 | 3.4 | 19.3 | 0.5 | 9.9 | 18 | 青海云杉 | 40 | 12.8 | 26.3 | 2.0 | 16.1 |
4 | 青海云杉 | 17 | 11.1 | 18.6 | 0.8 | 21.7 | 19 | 青海云杉 | 39 | 10.4 | 21.6 | 2.3 | 15.0 |
5 | 混交林 | 36 | 4.0 | 10.7 | 1.3 | 8.2 | 20 | 青海云杉 | 46 | 10.7 | 20.8 | 2.9 | 13.7 |
6 | 青海云杉 | 106 | 7.1 | 16.9 | 1.8 | 9.9 | 21 | 青海云杉 | 108 | 9.0 | 13.6 | 4.4 | 10.0 |
7 | 青海云杉 | 21 | 9.9 | 20.4 | 1.7 | 20.1 | 22 | 青海云杉 | 20 | 13.5 | 22.8 | 3.5 | 18.8 |
8 | 青海云杉 | 65 | 8.0 | 17.7 | 2.0 | 13.2 | 23 | 青海云杉 | 25 | 10.8 | 20.0 | 4.5 | 15.6 |
9 | 青海云杉 | 44 | 7.7 | 17.0 | 2.2 | 12.6 | 24 | 青海云杉 | 37 | 7.7 | 17.3 | 2.6 | 10.4 |
10 | 青海云杉 | 22 | 7.1 | 20.1 | 1.5 | 13.8 | 25 | 青海云杉 | 26 | 11.4 | 21.8 | 3.2 | 17.6 |
11 | 青海云杉 | 16 | 10.7 | 25.0 | 2.5 | 21.3 | 26 | 青海云杉 | 26 | 12.4 | 22.1 | 3.9 | 16.9 |
12 | 青海云杉 | 16 | 11.7 | 25.3 | 1.8 | 25.3 | 27 | 青海云杉 | 34 | 10.4 | 22.4 | 3.5 | 14.6 |
13 | 青海云杉 | 6 | 21.1 | 24.4 | 15.1 | 39.1 | 28 | 青海云杉 | 11 | 20.3 | 31.1 | 4.8 | 30.2 |
14 | 青海云杉 | 49 | 9.4 | 19.9 | 2.0 | 14.2 | 29 | 青海云杉 | 29 | 14.6 | 26.6 | 4.6 | 20.4 |
15 | 青海云杉 | 17 | 14.0 | 21.8 | 3.0 | 23.5 | 30 | 混交林 | 19 | 10.2 | 23.4 | 2.8 | 22.5 |
Fig. 1 Distribution of plots图1 样地分布图 |
Fig. 2 Compution units of HASM-AD algorithm图2 HASM-AD算法计算窗口 |
Fig. 3 Flow chart图3 单木树高提取算法流程图 |
Tab. 2 Parameter settings of comparison trials表2 对比试验参数设置 |
实验 | 分辨率(m) | 搜索半径(m) |
---|---|---|
1 | 0.1×0.1 | 0.5 |
2 | 0.25×0.25 | 0.5 |
3 | 0.5×0.5 | 0.5 |
4 | 0.5×0.5 | 1.0 |
5 | 0.5×0.5 | 1.5 |
Fig. 4 The influence of the parameter setting on tree vertex extracting图4 参数设置对树顶点提取的影响 |
Fig. 5 Tree height extraction results in Tianlaochi图5 天涝池树高识别提取结果 |
Fig. 6 Regression analysis between extracted tree height/number and measured tree height/number图6 提取树高和株数与实测值的回归分析 |
Tab. 3 Linear dependence between extracted tree height and measured tree height表3 提取树高与实测树高的线性回归关系 |
线性回归方程 | R2 | 说明 |
---|---|---|
y1=0.819x1+5.032 | 0.694 | y1、x1分别为提取的样地平均树高和实测值 |
y2=0.845x2+2.390 | 0.850 | y2、x2分别为提取的样地最大树高和实测值 |
y3=1.582x3+2.340 | 0.111 | y3、x3分别为提取的样地最小树高和实测值 |
y4=0.145x4+10.01 | 0.339 | y4、x4分别为提取的样地株数和实测值 |
Fig. 7 Simulation of CHM’s generating图7 CHM生成示意图 |
The authors have declared that no competing interests exist.
[1] |
|
[2] |
|
[3] |
|
[4] |
|
[5] |
|
[6] |
|
[7] |
|
[8] |
|
[9] |
|
[10] |
|
[11] |
|
[12] |
|
[13] |
|
[14] |
|
[15] |
|
[16] |
|
[17] |
|
[18] |
|
[19] |
|
[20] |
|
[21] |
|
[22] |
|
[23] |
|
/
〈 |
|
〉 |