一种基于无人机LiDAR和高空间分辨率卫星遥感数据的区域森林郁闭度估测方法：UnetR

徐恩恩; 郭颖; 陈尔学; 李增元; 赵磊; 刘清旺

doi:10.13203/j.whugis20210001

一种基于无人机LiDAR和高空间分辨率卫星遥感数据的区域森林郁闭度估测方法：UnetR

doi: 10.13203/j.whugis20210001

徐恩恩^1,2,
郭颖^2, ,,
陈尔学²,
李增元²,
赵磊²,
刘清旺²

1 西安科技大学, 陕西西安, 710054;
2 中国林业科学研究院资源信息研究所, 北京, 100091

基金项目:

国家重点研发计划（2017YFD0600900）

详细信息

作者简介:
徐恩恩,硕士生,研究方向为林业遥感应用。E-Mail:18404969811@163.com

A deep learning remote sensing estimation method (UnetR) for regional forest canopy closure combined with UAV LiDAR and high spatial resolution satellite remote sensing data

1 Xi' an University of Science and Technology, Xi' an 710054, China;
2 Research Institute of Forest Resources Information Technique, Chinese Academy of Forestry, Beijing 100091, China

Funds:

the National Key Research and Development Program of China (2017YFD0600900).

摘要: 森林郁闭度是森林资源调查中的一个重要因子，对森林质量评价具有重要的作用。随着人工智能技术和遥感技术的不断发展，研究如何利用深度学习有效协同应用不同空间覆盖能力的遥感数据实现区域森林郁闭度的估测具有重要意义。本研究提出了一种协同应用高密度无人机激光雷达和高空间分辨率卫星遥感数据，对区域森林郁闭度进行定量估测的深度学习模型（UnetR）。通过对用于图像分类的Unet的损失函数进行改进，并在卷积层后加入批量归一化层，使其具有对连续变量进行定量估测的能力。与全卷积神经网络（fully convolutional networks，FCN）、随机森林（random forest，RF）和支持向量机回归（supportvector regression，SVR）模型的对比评价结果表明，UnetR模型的均方根误差较低，估测精度较高，为实现区域森林郁闭度遥感监测提供了一种人力成本低、自动化程度高的估测方法。
- 区域森林郁闭度 /
- 深度学习 /
- Unet /
- 光学遥感 /
- 激光雷达
Abstract: Forest canopy density is an important factor in forest resource surveys. It plays an important role in forest quality evaluation and forest resource management. In recent years, artificial intelligence technology and remote sensing technology have developed rapidly and they have been successfully applied to forestry remote sensing quantitative estimation. There is a great significance to study how to use deep learning methods to effectively integrate the remote sensing data with different spatial coverage capabilities in regional forest canopy closure estimation. Therefore, this research proposed a deep learning model (UnetR) which aimed to estimate forest canopy closure based on deep learning model with the high-density light detection and ranging and high spatial resolution satellite imagery. By optimizing the loss function of Unet for image classification, and adding a batch normalization layer after the convolution layer, the model had the ability to quantitatively estimate continuous variables. The comparative evaluation results with fully convolutional networks (FCN), random forest (RF) and support vector regression (SVR) models showed that the root mean square error of the UnetR model was lower, the estimation accuracy was higher (determination coefficient (R²)=0.777, root mean square error (RMSE)=0.137, estimation accuracy (EA)=75.60%). This research provided a low labor cost and high degree of automation estimation method for remote sensing monitoring of regional forest canopy closure.
- regional forest canopy closure /
- deep learning /
- Unet /
- optical remote sensing /
- LiDAR

[1]	Liu Q, Fu L, Wang G, et al. Improving Estimation of Forest Canopy Cover by Introducing Loss Ratio of Laser Pulses Using Airborne LiDAR[J]. IEEE Transactions on Geoence and Remote Sensing, 2019, 58(1):567-585
[2]	Stojanova D, Panov P, Gjorgjioski V, et al. Estimating vegetation height and canopy cover from remotely sensed data with machine learning[J]. Ecological Informatics, 2010, 5(4):256-266
[3]	Ahmed O S, Franklin S E, Wulder M A. Integration of lidar and landsat data to estimate forest canopy cover in coastal British Columbia[J]. Photogrammetric Engineering & Remote Sensing, 2014, 80(10):953-961
[4]	Korhonen L, Ali-Sisto D, Tokola T. Tropical forest canopy cover estimation using satellite imagery and airborne lidar reference data[J]. Silva Fenn, 2015, 49(5):1-18
[5]	Elias A, Daniel H. The Use of Three-Dimensional Convolutional Neural Networks to Interpret LiDAR for Forest Inventory[J]. Remote Sensing, 2018, 10(4):649.
[6]	Dong L, Du H, Han N, et al. Application of Convolutional Neural Network on Lei Bamboo Above-Ground-Biomass (AGB) Estimation Using Worldview-2[J]. Remote Sensing, 2020, 12(6):958
[7]	Shah S A A, Manzoor M A, Bais A. Canopy Height Estimation at Landsat Resolution Using Convolutional Neural Networks[J]. Machine Learning and Knowledge Extraction, 2020, 2(1):23-36
[8]	Martin L, Andrey K, Marjan A, et al. Classification and Segmentation of Satellite Orthoimagery Using Convolutional Neural Networks[J]. Remote Sensing, 2016, 8(4):329-329
[9]	Long J, Shelhamer E, Darrell T. Fully convolutional networks for semantic segmentation[C]//Proceedings of the IEEE conference on computer vision and pattern recognition. 2015:3431-3440
[10]	Cao, K, Zhang, X. An Improved Res-UNet Model for Tree Species Classification Using Airborne High-Resolution Images[J]. Remote Sensing, 2020, 12(7):1128
[11]	Pan, Z, Xu, J, Guo Y, Hu Y, et al. Deep Learning Segmentation and Classification for Urban Village Using a Worldview Satellite Image Based on U-Net[J]. Remote Sensing, 2020, 12(10):1574
[12]	Ioffe S, Szegedy C. Batch normalization:Accelerating deep network training by reducing internal covariate shift[J]. Computer Science, 2015
[13]	Draper N R, Smith H. Applied regression analysis[M]. Wiley Series in Probability and Statistics, 1998
[14]	Clark. Basic Statistics[M]//Handbook of Pulp. John Wiley & Sons, Ltd, 2008
[15]	BREIMAN L. Using iterated bagging to Debias regressions[J]. Machine Learning, 2001, 45(3):261-277
[16]	JOG A, CARASS A, ROY S, et al. Random forest regression for magnetic resonance image synthesis[J]. Medical Image Analysis, 2017, 35:475-488
[17]	WANG C Y, YU Y C. Growing season variations of soil detachment under wheatgrass and switchgrass lands in the loess area of China[J]. Chinese Journal of Soil Science, 2016, 47(4):790-796
[18]	CHEN Y, XU P, CHU Y, et al. Short-term electrical load forecasting using the support vector regression (SVR) model to calculate the demand response baseline for office buildings[J]. Applied Energy, 2017, 195:659-670
[19]	Li J, Mao X. Comparison of Canopy Closure Estimation of Plantations Using Parametric, Semi-Parametric, and Non-Parametric Models Based on GF-1 Remote Sensing Images[J]. Forests, 2020, 11(5):597

[1]	王哲, 赵超英, 刘晓杰, 李滨. 西藏易贡滑坡演化光学遥感分析与InSAR形变监测 . 武汉大学学报 ● 信息科学版, doi: 10.13203/j.whugis20210168
[2]	曹伟, 陈动, 史玉峰, 曹震, 夏少波. 激光雷达点云树木建模研究进展与展望 . 武汉大学学报 ● 信息科学版, doi: 10.13203/j.whugis20190275
[3]	陈行, 罗斌. 利用动态上采样滤波深度网络进行多角度遥感影像超分辨率重建 . 武汉大学学报 ● 信息科学版, doi: 10.13203/j.whugis20200651
[4]	郭旦怀, 张鸣珂, 贾楠, 王彦棡. 融合深度学习技术的用户兴趣点推荐研究综述 . 武汉大学学报 ● 信息科学版, doi: 10.13203/j.whugis20200334
[5]	巨袁臻, 许强, 金时超, 李为乐, 董秀军, 郭庆华. 使用深度学习方法实现黄土滑坡自动识别 . 武汉大学学报 ● 信息科学版, doi: 10.13203/j.whugis20200132
[6]	李鹏, 普思寻, 李振洪, 王厚杰. 2000年以来胶州湾海岸线光学与SAR多源遥感变化监测研究 . 武汉大学学报 ● 信息科学版, doi: 10.13203/j.whugis20180483
[7]	陆会燕, 李为乐, 许强, 董秀军, 代聪, 王栋. 光学遥感与InSAR结合的金沙江白格滑坡上下游滑坡隐患早期识别 . 武汉大学学报 ● 信息科学版, doi: 10.13203/j.whugis20190086
[8]	潘银, 邵振峰, 程涛, 贺蔚. 利用深度学习模型进行城市内涝影响分析 . 武汉大学学报 ● 信息科学版, doi: 10.13203/j.whugis20170217
[9]	赵文豪, 闫利, 王成义, 马昕, 李松, 马跃. 基于三维大气探测激光雷达的大气颗粒物污染探测 . 武汉大学学报 ● 信息科学版, doi: 10.13203/j.whugis20190105
[10]	张兵. 遥感大数据时代与智能信息提取 . 武汉大学学报 ● 信息科学版, doi: 10.13203/j.whugis20180172
[11]	张继贤, 段敏燕, 林祥国, 臧艺. 激光雷达点云电力线三维重建模型的对比与分析 . 武汉大学学报 ● 信息科学版, doi: 10.13203/j.whugis20150385
[12]	樊恒, 徐俊, 邓勇, 向金海. 基于深度学习的人体行为识别 . 武汉大学学报 ● 信息科学版, doi: 10.13203/j.whugis20140110
[13]	李建, 陈晓玲, 刘子潇, 田礼乔. HJ-1光学遥感影像船只检测——以鄱阳湖为例 . 武汉大学学报 ● 信息科学版, doi: 10.13203/j.whugis20130095
[14]	史硕, 龚威, 祝波, 宋沙磊. 新型对地观测多光谱激光雷达及其控制实现 . 武汉大学学报 ● 信息科学版,
[15]	吕立蕾, 龚威, 宋沙磊, 祝波. 地物反射率探测激光雷达回波信号的小波去噪 . 武汉大学学报 ● 信息科学版,
[16]	邢艳秋, 王立海. 基于ICESat-GLAS完整波形的坡地森林冠层高度反演研究——以吉林长白山林区为例 . 武汉大学学报 ● 信息科学版,
[17]	曾齐红, 毛建华, 李先华, 刘学锋. 激光雷达点云平面拟合过滤算法 . 武汉大学学报 ● 信息科学版,
[18]	赖旭东, 万幼川. 一种针对激光雷达强度图像的滤波算法研究 . 武汉大学学报 ● 信息科学版,
[19]	李清泉, 李必军, 陈静. 激光雷达测量技术及其应用研究 . 武汉大学学报 ● 信息科学版,
[20]	邵振峰, 孙悦鸣, 席江波, 李岩. 智能优化学习的高空间分辨率遥感影像语义分割 . 武汉大学学报 ● 信息科学版, doi: 10.13203/j.whugis20200640

点击查看大图

计量

文章访问数: 92
HTML全文浏览量: 27
PDF下载量: 13
被引次数: 0

姓名
邮箱
手机号码
标题
留言内容
验证码

留言板

一种基于无人机LiDAR和高空间分辨率卫星遥感数据的区域森林郁闭度估测方法：UnetR

doi: 10.13203/j.whugis20210001

作者简介:
徐恩恩,硕士生,研究方向为林业遥感应用。E-Mail:18404969811@163.com

A deep learning remote sensing estimation method (UnetR) for regional forest canopy closure combined with UAV LiDAR and high spatial resolution satellite remote sensing data

计量

一种基于无人机LiDAR和高空间分辨率卫星遥感数据的区域森林郁闭度估测方法：UnetR

doi: 10.13203/j.whugis20210001

1 西安科技大学, 陕西西安, 710054;

2 中国林业科学研究院资源信息研究所, 北京, 100091

作者简介:
徐恩恩,硕士生,研究方向为林业遥感应用。E-Mail:18404969811@163.com

English Abstract

A deep learning remote sensing estimation method (UnetR) for regional forest canopy closure combined with UAV LiDAR and high spatial resolution satellite remote sensing data

1 Xi' an University of Science and Technology, Xi' an 710054, China;

2 Research Institute of Forest Resources Information Technique, Chinese Academy of Forestry, Beijing 100091, China

全文HTML

目录

留言板

一种基于无人机LiDAR和高空间分辨率卫星遥感数据的区域森林郁闭度估测方法：UnetR

doi: 10.13203/j.whugis20210001

作者简介: 徐恩恩,硕士生,研究方向为林业遥感应用。E-Mail:18404969811@163.com

A deep learning remote sensing estimation method (UnetR) for regional forest canopy closure combined with UAV LiDAR and high spatial resolution satellite remote sensing data

计量

出版历程

一种基于无人机LiDAR和高空间分辨率卫星遥感数据的区域森林郁闭度估测方法：UnetR

doi: 10.13203/j.whugis20210001

1 西安科技大学, 陕西 西安, 710054; 2 中国林业科学研究院资源信息研究所, 北京, 100091

作者简介: 徐恩恩,硕士生,研究方向为林业遥感应用。E-Mail:18404969811@163.com

English Abstract

A deep learning remote sensing estimation method (UnetR) for regional forest canopy closure combined with UAV LiDAR and high spatial resolution satellite remote sensing data

1 Xi' an University of Science and Technology, Xi' an 710054, China; 2 Research Institute of Forest Resources Information Technique, Chinese Academy of Forestry, Beijing 100091, China

全文HTML

目录

作者简介:
徐恩恩,硕士生,研究方向为林业遥感应用。E-Mail:18404969811@163.com

1 西安科技大学, 陕西西安, 710054;

2 中国林业科学研究院资源信息研究所, 北京, 100091

作者简介:
徐恩恩,硕士生,研究方向为林业遥感应用。E-Mail:18404969811@163.com

1 Xi' an University of Science and Technology, Xi' an 710054, China;

2 Research Institute of Forest Resources Information Technique, Chinese Academy of Forestry, Beijing 100091, China