基于XGBoost算法的近紫外通道地表反射率模拟

doi:10.13474/j.cnki.11-2246.2023.0170

测绘通报 ›› 2023, Vol. 0 ›› Issue (6): 68-74,103.doi: 10.13474/j.cnki.11-2246.2023.0170

基于XGBoost算法的近紫外通道地表反射率模拟

奥勇¹, 李红丽^2,3, 张文娟³, 秦梦^2,3

1. 长安大学土地工程学院, 陕西西安 710054;
2. 长安大学地球科学与资源学院, 陕西西安 710054;
3. 中国科学院空天信息创新研究院, 北京 100094

收稿日期:2022-11-10 修回日期:2023-04-20 发布日期:2023-07-05
通讯作者: 张文娟。E-mail:zhangwj@aircas.ac.cn
作者简介:奥勇(1965-),男,博士,副教授,研究方向为遥感科学与技术、地理信息系统。E-mail:aoyong@chd.edu.cn
基金资助:
中国科学院青年创新促进会资助项目(2019132)

Near-ultraviolet channel surface reflectance simulation based on XGBoost algorithm

AO Yong¹, LI Hongli^2,3, ZHANG Wenjuan³, QIN Meng^2,3

1. School of Land Engineering, Chang'An University, Xi'an 710054, China;
2. School of Earth Science and Resources, Chang'An University, Xi'an 710054, China;
3. Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China

Received:2022-11-10 Revised:2023-04-20 Published:2023-07-05

摘要/Abstract

摘要： 紫外谱段在全球极光探测、海洋溢油、大气辉光等领域具有重要应用价值,其地表反射特性是研究中的重要背景数据,但现有卫星数据资源较少,难以满足应用需求。针对这一问题,本文提出了基于机器学习XGBoost算法的近紫外(350～400 nm)通道地表反射率数据模拟方法。首先,选取Sentinel-2 MSI 2、3、4通道多光谱数据为数据源,结合其通道特点基于USGS地物光谱数据库获取植被、水体、土壤等典型地物光谱数据,并等效计算到相应通道。其次,对数据源和待模拟通道开展相关性分析,Sentinel-2 MSI 2、3、4通道与待模拟通道相关系数均大于0.88,这表明基于该数据源可开展近紫外地表反射率数据模拟。然后,基于等效计算后的典型地物光谱数据集,利用XGBoost算法构建近紫外通道地表反射率回归模型。精度分析结果表明,所有通道模型决定系数(R²)均达到0.91以上,均方根误差(RMSE)均小于0.076,平均绝对误差百分比(MAPE)整体在20%以内,且上述3个精度指标针对不同类别样本的标准差在0.021 2范围内,可见模型精度较高,同时具有良好的稳健性。最后,基于Sentinel-2 MSI 2、3、4通道图像数据,生成355、365、375、385、395 nm的地表反射率模拟图像,图像较好地体现了地物光谱特性。

关键词: 近紫外, 地表反射率模拟, 机器学习, XGBoost

Abstract: The ultraviolet spectrum has significant applications in the fields of global auroral detection,marine oil spill,atmospheric glow,etc. Surface reflectance is important background data in the research. However,the existing satellite data resources are relatively insufficient to meet the application needs. In this study,a machine learning-based on XGBoost algorithm is proposed for simulating surface reflectance data in the near-ultraviolet (N-UV) (350～400 nm) spectral channel. Firstly,Sentinel-2 MSI 2,3 and 4 channels are selected as the data source and the spectral of vegetation,water,soil and other typical features are extract based on the USGS spectral database,then equivalently calculated to the corresponding channels. Secondly,the correlation analysis between the data source and the channel to be simulated is carried out. The correlation coefficients between Sentinel-2 MSI 2,3 and 4 channels and the channels to be simulated are all greater than 0.88,which indicates that the N-UV surface reflectance simulation can be carried out based on this data source. Thirdly,based on the typical spectral data set after the equivalent calculation construct XGBoost regression model to simulate the N-UV channel surface reflectance. Results indicate that the coefficient of determination (R²) of all the channel models is above 0.91,the root mean square error (RMSE) is less than 0.076,the mean absolute error percentage (MAPE) is within 20%,and the standard deviation of the above three accuracy indicators for different categories of samples is within 0.021 2,which shows that the model has high accuracy and robustness. Finally,based on the Sentinel-2 MSI 2,3 and 4 channels image data,the simulated images of surface reflectance at 355,365,375,385 and 395 nm are generated,and the images better reflect the spectral characteristics of the surface.

Key words: near-ultraviolet, surface reflectance simulation, machine learning, XGBoost

中图分类号:

P237

奥勇, 李红丽, 张文娟, 秦梦. 基于XGBoost算法的近紫外通道地表反射率模拟[J]. 测绘通报, 2023, 0(6): 68-74,103.

AO Yong, LI Hongli, ZHANG Wenjuan, QIN Meng. Near-ultraviolet channel surface reflectance simulation based on XGBoost algorithm[J]. Bulletin of Surveying and Mapping, 2023, 0(6): 68-74,103.

参考文献

[1] 童庆禧,孟庆岩,杨杭. 遥感技术发展历程与未来展望[J]. 城市与减灾,2018(6):2-11.
[2] 张艺蔚,陶邦一,毛志华,等. 海洋水色卫星紫外波段的偏振特性分析[J]. 光学学报,2020,40(6):1-11.
[3] 方四安,黄小仙,尹达一,等. 海洋溢油模拟目标的紫外反射特性研究[J]. 光谱学与光谱分析,2010,30(3):738-742.
[4] 赵冉,胡启后,孙中平,等. 天地一体化遥感监测大气污染技术进展[J]. 环境科学研究,2021,34(1):28-40.
[5] 王淑荣,李福田,曲艺. 空间紫外光学遥感技术与发展趋势[J]. 中国光学与应用光学,2009(1):17-22.
[6] 张建勇,钟生东. 紫外线技术在军事工程技术中的应用[J]. 光学技术,2000,26(4):308-312.
[7] 赵少华,杨晓钰,李正强,等. 臭氧卫星遥感六十年进展[J]. 遥感学报,2022,26(5):817-833.
[8] 王丽丽,赵鸿志,张可立. 海洋水色卫星的发展现状及趋势[J]. 航天器工程,2021,30(6):44-51.
[9] 童庆禧,张兵,张立福. 中国高光谱遥感的前沿进展[J]. 遥感学报,2016,20(5):689-707.
[10] JACQUEMOUD S,BARET F. PROSPECT:a model of leaf optical properties spectra[J]. Remote Sensing of Environment,1990,34(2):75-91.
[11] 刘瑶,张文娟,张兵,等. 基于光谱混合的大气吸收通道发射率图像模拟[J]. 遥感技术与应用,2017,32(4):674-682.
[12] LIU S,ZHOU J,QIU Y,et al. The FIRST model:spatiotemporal fusion incorrporting spectral autocorrelation[J]. Remote Sensing of Environment,2022,279:113111.
[13] 杨倩倩,靳才溢,李同文,等.数据驱动的定量遥感研究进展与挑战[J].遥感学报,2022,26(2):268-285.
[14] CHEN Tianqi,GUESTRIN C. XGBoost:a scalable tree boosting system[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Francisco:ACM Press,2016:785-794.
[15] LIU H,LI Q,BAI Y,et al. Improving satellite retrieval of oceanic particulate organic carbon concentrations using machine learning methods[J]. Remote Sensing of Environment,2021,256:112316.
[16] MULTI Y,SUN W,CHI Y,et al. Machine learning-based retrieval of day and night cloud macrophysical parameters over East Asia using Himawari-8 data[J]. Remote Sensing of Environment,2022,273:112971.
[17] GREEN R O,PAVRI B E,CHRIEN T G. On-orbit radiometric and spectral calibration characteristics of EO-1 Hyperion derived with an underflight of AVIRIS and in situ measurements at Salar de Arizaro,Argentina[J]. IEEE Transactions on Geoscience and Remote Sensing,2003,41(6):1194-1203.
[18] TAYLOR R. Interpretation of the correlation coefficient:a basic review[J]. Journal of Diagnostic Medical Sonography,1990,6(1):35-39.
[19] PEDREGOSA F,VAROQUAUX G,GRAMFORT A,et al. Scikit-learn:machine learning in python[J]. Journal of machine Learning research,2011,12:2825-2830.
[20] BERGSTRA J,KOMER B,ELIASMITH C,et al. Hyperopt:a Python library for model selection and hyperparameter optimization[J]. Computational Science & Discovery,2015,8(1):014008.

基于XGBoost算法的近紫外通道地表反射率模拟

Near-ultraviolet channel surface reflectance simulation based on XGBoost algorithm

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	李鹏, 周虹利, 林是聪, 王厚杰, 李振洪. 海洋塑料垃圾卫星遥感探测方法[J]. 测绘通报, 2023, 0(6): 20-26.
[2]	刘明月, 郑浩, 陈星彤, 杨晓芜, 宋敬茹, 张永彬, 满卫东. GEE支持下联合植被物候特征与机器学习的入侵植物互花米草提取[J]. 测绘通报, 2023, 0(6): 36-43.
[3]	孔瑞瑶, 谢涛, 马明, 孔瑞林. CatBoost模型在水深反演中的应用[J]. 测绘通报, 2022, 0(7): 33-37.
[4]	温雨笑, 吕杰, 马庆勋, 张鹏, 徐汝岭. 高光谱和LiDAR联合反演森林生物量研究[J]. 测绘通报, 2022, 0(7): 38-42.
[5]	张驰, 白志辉, 李亮, 陈冉丽, 吴侃. 利用三维激光扫描技术监测矿区桥梁结构形变[J]. 测绘通报, 2022, 0(4): 122-129.
[6]	王刚, 丁华祥. Sentinel-2A数据支持下的雷州半岛植被类型识别[J]. 测绘通报, 2022, 0(3): 76-82.
[7]	吕蓓茹, 彭玲, 李樵民. 顾及样本敏感性的滑坡易发性评价[J]. 测绘通报, 2022, 0(11): 20-25.
[8]	张自强, 刘涛, 杜萍, 锁旭宏, 杨国林. 图残差神经网络支持下的建筑物群组模式分类[J]. 测绘通报, 2022, 0(1): 1-7.
[9]	叶烨星, 杨飞. 基于介数中心性的交通拥堵指数计算[J]. 测绘通报, 2021, 0(5): 86-90.
[10]	张鹏, 马庆勋, 吕杰, 季金亮, 李紫微. 机器学习算法在森林地上生物量估算中的应用[J]. 测绘通报, 2021, 0(12): 28-32.
[11]	夏汉庸, 尹和军, 徐教煌, 王嘉伟, 黄毅. 基于机器学习的多施工参数盾构施工姿态预测[J]. 测绘通报, 2021, 0(1): 157-160,164.
[12]	王晓静, 唐超, 杨晓飞. 激光点云在地铁盾构隧道病害诊断中的应用[J]. 测绘通报, 2020, 0(9): 33-37.
[13]	李月, 何宏昌, 王晓飞, 张国民. 农作物冠层光谱分析及反演技术综述[J]. 测绘通报, 2019, 0(9): 13-17.
[14]	修晓敏, 周淑芳, 陈黔, 蒙继华, 董文全, 杨广斌, 李晓松. 基于Google Earth Engine与机器学习的省级尺度零散分布草地生物量估算[J]. 测绘通报, 2019, 0(3): 46-52,75.
[15]	张金盈, 姚光虎, 林琳, 郭怀轩. 结合主动学习和词袋模型的高分二号遥感影像自动化分类[J]. 测绘通报, 2019, 0(2): 103-107.