Please wait a minute...
img

官方微信

遥感技术与应用  2022, Vol. 37 Issue (2): 333-341    DOI: 10.11873/j.issn.1004-0323.2022.2.0333
LUCC专栏     
基于样本迁移的干旱区地表覆盖快速更新
盖一铭1,2,3(),阿里木·赛买提1,2,3,王伟1,2,3,吉力力·阿不都外力1,2,3()
1.中国科学院新疆生态与地理研究所 荒漠与绿洲生态国家重点实验室,新疆 乌鲁木齐 830011
2.中国科学院大学,北京 10049
3.中国科学院中亚生态环境研究中心,新疆 乌鲁木齐 830011
Sample Transferring based Fast Land Cover Updating in Arid Land
Yiming Gai1,2,3(),Samat Alim1,2,3,Wei Wang1,2,3,Abuduwaili Jilili1,2,3()
1.State Key Laboratory of Desert and Oasis Ecology,Xinjiang Institute of Ecology and Geography,Chinese Academy of Sciences,Urumqi 830011,China
2.University of Chinese Academy of Sciences,Beijing 10049,China
3.Chinese Academy of Sciences Research Center for Ecology and Environment of Central Asia,Urumqi 830011,China
 全文: PDF(3772 KB)   HTML
摘要:

阿姆河三角洲作为典型干旱区,干旱胁迫和次生的盐胁迫决定了本地区生态环境的复杂性和独特性,给遥感地表覆盖制图带来一定的困难。在土地利用/覆盖(LULC)遥感图像分类任务中,数量大、质量高、成本低的样本和速度快、性能稳定的分类器是高效实现高精度分类的关键。在一些偏远地区开展土地利用/地表覆盖遥感图像分类依然面临着标记样本空间上稀疏、时间上不连续甚至是缺失,人工收集成本高等问题。为此,结合最优树集成和样本迁移的思想,构建了一种高效的地表覆盖自动更新的新方法。该方法通过变化检测在历史产品上的同期影像上进行样本标签的标记,并将过去的地表覆盖类型标签转移到同源目标影像上,使用最优树集成(Ensemble of optimum trees, OTE)完成地表覆盖自动分类。根据阿姆河三角洲地区地表覆盖分类试验结果,表明该方法可以提取有效的地表覆盖标签,并能较高精度发实现土地利用/地表覆盖的自动分类更新。

关键词: 样本迁移最优树集成变化检测地表覆盖变化遥感图像分类干旱区地表覆盖    
Abstract:

Amu river delta, as a typical arid land, was threatened by drought and salination, which contribute to the complexity and specificality of its ecological environment. In the Land Use/Land Cover (LULC) Remote Sensing (RS) image classification tasks, collecting large number of high quality samples at low-cost and a high efficient and robust classifier are always the crucial factors to obtain high-accuracy classification results. However, it was still problems facing RS imageries classification in some remote areas that marked samples were sparsely distributed, timely dissected or even intermittent, and manual tasks for field sampling cost high. In this end, a new frame of automatic land cover classification based on ensemble of optimum trees and sample transfer was promoted in this paper. In this frame, sample labels were marked on the historical image which is same time and source with the product, then these labels were transferred into targeted RS image. Then, OTE method classification was performed. According to the results in this paper, the OTE with sample transferring based method can extract land cover labels efficiently and update LULC map in a fine accuracy.

Key words: Sample transfer    Ensemble of optimum trees    Change detection    Land cover change    RS image classification    Arid land cover
收稿日期: 2021-01-23 出版日期: 2022-06-17
ZTFLH:  TP79  
基金资助: 国家自然科学基金项目“样本与特征迁移的中亚典型城市覆被精细分类方法研究”(42071424);中国科学院战略性先导专项“咸海退缩产生的盐尘及其环境影响”(XDA2006030102);中国科学院青年创新促进会(2018476)
通讯作者: 吉力力·阿不都外力     E-mail: gaiyiming18@mails.ucas.ac.cn;jilil@ms.xjb.ac.cn
作者简介: 盖一铭(1995-),男,山东郯城人,硕士研究生,主要从事环境演变与气候变化研究。E?mail: gaiyiming18@mails.ucas.ac.cn
服务  
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章  
盖一铭
阿里木·赛买提
王伟
吉力力·阿不都外力

引用本文:

盖一铭,阿里木·赛买提,王伟,吉力力·阿不都外力. 基于样本迁移的干旱区地表覆盖快速更新[J]. 遥感技术与应用, 2022, 37(2): 333-341.

Yiming Gai,Samat Alim,Wei Wang,Abuduwaili Jilili. Sample Transferring based Fast Land Cover Updating in Arid Land. Remote Sensing Technology and Application, 2022, 37(2): 333-341.

链接本文:

http://www.rsta.ac.cn/CN/10.11873/j.issn.1004-0323.2022.2.0333        http://www.rsta.ac.cn/CN/Y2022/V37/I2/333

图1  研究区示意图
图2  样本迁移的遥感影像分类流程图
类别分类精度/%
SVMRFXGBExTOTE
耕地76.5077.8276.3080.2477.38
森林93.3094.0494.3596.3295.76
草地73.9978.2472.5880.2679.81
灌丛87.7090.2889.1790.3290.82
湿地75.8277.3378.5682.5484.12
水体95.1596.0996.2396.5095.86
裸地95.8796.8096.5296.3497.08
建筑85.8885.4487.0686.7789.42
OA85.4687.0786.2988.7888.88
Kappa0.830.850.840.870.87
表1  5种分类器总体精度
图3  分类结果对比(a为遥感影像,b为产品,c为OTE分类结果)
图4  各分类器训练时间对比图
地类耕地森林草地灌丛湿地水体裸地建筑
百分比/%16.560.00714.130.220.902.8065.190.13
表 2  各个地表覆盖类型的面积占比
1 Grimm N B, Faeth S H, Golubiewski N E, et al. Global change and the ecology of cities[J].Science,2008,319(5864): 756-60. DOI: 10.1126/science.1150195 .
doi: 10.1126/science.1150195
2 Fischer M, Bossdorf O, Gockel S, et al. Implementing large-scale and long-term functional biodiversity research: The Biodiversity Exploratories[J]. Basic and Applied Ecology, 2010, 11(6): 473-485.
3 Stehman S V, Foody G M. Key issues in rigorous accuracy assessment of land cover products[J/OL]. Remote Sensing of Environment,2019,23123..
4 Du P J, Bai X, Tan K, et al. Advances of four machine larning methods for spatial data handling: A review[J/OL]. Journal of Geovisualization and Spatial Analysis, 2020, 4(1): https:∥DOI.org/10.1007/s41651-020-00048-5.
5 Du Peijun, Xia Junshi, Xue Zhaohui, et al. Review of hyperspectral remote sensing image classification[J]. Journal of Remote Sensing, 2016, 20(2): 236-256.
5 杜培军, 夏俊士, 薛朝辉. 高光谱遥感影像分类研究进展[J]. 遥感学报, 2016, 20(2): 236-256.
6 Wu Tianjun, Luo Jiancheng, Xia Liegang, et al. An automatic sample collection method for object-oriented classification of remotely sensed imageries based on transfer learning[J]. Acta Geodetica et Cartographica Sinica,2014,43(9):908-916.
6 吴田军,骆剑承,夏列钢.迁移学习支持下的遥感影像对象级分类样本自动选择方法[J].测绘学报,2014,43(9): 908-916.
7 Lin Cong, Li Erzhu, Du Peijun. An automatic approach for remote sensing classification supported by Sample Transfer[J]. Bulletin of Surveying and Mapping, 2018(4): 57-62.
7 林聪, 李二珠, 杜培军. 样本迁移支持下的遥感影像自动分类方法[J]. 测绘通报, 2018(4):57-62.
8 Liu Wanjun, Li Tianhui, Qu Haicheng. Hyperspectral similar sample classification algorithm based on fisher criterion and TrAdaboost[J].Remote Sensing for Land & Resources,2018,30(4):41-48.
8 刘万军, 李天慧, 曲海成. 基于Fisher准则和TrAdaboost的高光谱相似样本分类算法[J]. 国土资源遥感, 2018,30(4):41-48.
9 Pan H, Tong X, Xu X, et al. Updating of land cover maps and change analysis using GlobeLand30 product: A case study in Shanghai metropolitan area,China[J]. Remote Sensing,2020,12(19): https:∥DOI.org/10.3390/rs12193147.
10 Song A, Choi J. Fully convolutional networks with multiscale 3D filters and Transfer Learning for change detection in high spatial resolution satellite images[J]. Remote Sensing, 2020, 12(5). .
11 De Lima R P, Marfurt K. Convolutional neural Network for remote sensing scene classification: Transfer Learning Analysis[J]. Remote Sensing,2020,12(1):20. .
12 Gu Chunxia, Li Dacheng. Application of spatio-temporal fusion technology in time series classification of regional land cover[J]. Bulletin of Surveying and Mapping, 2021(5): 20-24.
12 古春霞,李大成.时空融合技术在区域地表覆盖时序分类中的应用[J].测绘通报,2021(5):20-24.
13 Li Binbin, Xie Huan, Tong Xiaohua, et al. Land cover classification using ICESat-2 data with random forest[J]. Infrared and Laser Engineering, 2020, 49(11): 115-121.
13 李彬彬, 谢欢, 童小华, 等. 基于随机森林的ICESat-2卫星数据地表覆盖分类[J]. 红外与激光工程, 2020, 49(11): 115-121.
14 Li Bing, Lu Xiaoping, Li Xinshe, et al. Extraction method of wetland surface factors using GF-2 images on optimal features[J]. Geomatics & Sptial Information Technolog, 2018,41(9):49-52.
14 李冰,卢小平,李新社,等.特征优选的GF-2影像湿地地表覆盖要素提取[J].测绘与空间地理信息,2018,41(9):49-52.
15 Oshiro T M, Perez P S, Baranauskas J A. How many trees in a random forest[C]∥ Proceedings of the Machine Learning and Data Mining in Pattern Recognition, F, 2012.
16 Bernard S, Heutte L, Adam S. On the selection of decision trees in Random Forests[C]∥ Proceedings of the International Joint Conference on Neural Network, F, 2009.
17 Abdelali Z, Hicham M, Abdelwahed N. An Ensemble of Optimal Trees for Software Development Effort Estimation[C]∥ Proceedings of the International Conference Advanced Information Technology, Services and Systems, F, 2018.
18 Samat A, Li E Z, Wang W, et al. Meta-XGBoost for hyperspectral image classification using Extended MSER-Guided morphological profiles[J]. Remote Sensing, 2020, 12(12): 23. DOI: 10.3390/rs12121973 .
doi: 10.3390/rs12121973
19 Zhang H, Wang M. Search for the smallest random forest[J]. Statistics and Its Interface,2009,2(3):381-388.DOI:10. 1093/biostatistics/kxaa021 .
doi: 10. 1093/biostatistics/kxaa021
20 Adler W, Gefeller O, Gul A, et al. Ensemble pruning for glaucoma detection in an unbalanced data set[J]. Methods of Information in Medicine, 2016, 55(6): 557-63. DOI: 10.3414/ME16-01-0055 .
doi: 10.3414/ME16-01-0055
21 Khan Z, Gul A, Perperoglou A, et al. Ensemble of optimal trees, random forest and random projection ensemble classification[J]. Advanced Data Analysis and Classification, 2020, 14(1): 97-116. DOI: 10.1007/s11634-019-00364-9 .
doi: 10.1007/s11634-019-00364-9
22 Mo Guifen, Feng Jianzhong, Wang Zhongmei, et al. Spatial-temporal evolution characteristics of landscape ecological risk in the transboundary basin of Amu Darya River, Central Asia[J]. Agricultural Research in the Arid Areas, 2022,40(1):123-131.
22 莫贵芬,冯建中,王中美,等.中亚阿姆河跨境流域景观生态风险时空演变特征分析[J].干旱地区农业研究,2022,40(1):123-131.
23 Hu Y, Duan W, Chen Y, et al. An integrated assessment of runoff dynamics in the Amu Darya River Basin: Confronting climate change and multiple human activities,1960-2017[J]. Journal of Hydrology, 2021, 603: 126905. .
24 He H, Hamdi R, Cai P, et al. Impacts of historical land use/cover change (1980–2015) on summer climate in the Aral Sea Region[J]. Journal of Geophysical Research: Atmospheres, 2021, 126(6)
25 Han Q F, Luo G P, Li C F, et al. Response of carbon dynamics to climate change varied among different vegetation types in Central Asia[J]. Sustainability, 2018, 10(9): 15. DOI:10.3390/su10093288 .
doi: 10.3390/su10093288
26 Li Qi, Li Fadong, Wang Guoqin, et al. Development of irrigated agriculture in Uzbekistan an its impact on ecological enviroment and economic development[J]. Arid Land Geography, 2021,44(6):1810-1820.
26 李琦,李发东,王国勤,等.乌兹别克斯坦灌溉农业发展及其对生态环境和经济发展的影响[J].干旱区地理,2021,44(6):1810-1820.
27 Ye H, Huang X T, Luo G P, et al. Improving remote sensing-based net primary production estimation in the grazed land with defoliation formulation model[J]. Journal of Mountain Science, 2019, 16(2): 323-336. .
28 Roy D P, Wulder M A, Lovehand T R, et al. Landsat 8: Science and product vision for terrestrial global change research[J]. Remote Sensing of Environment, 2014,145:154-172. DOI:10.1016/j.rse.2014.02.001 .
doi: 10.1016/j.rse.2014.02.001
29 Zhu Z, Woodcock C E. Continuous change detection and classification of land cover using all available Landsat data[J]. Remote Sensing of Environment,2014,144:152-171. DOI:10. 1016/j.rse.2014.01.011 .
doi: 10. 1016/j.rse.2014.01.011
30 Zhang Xiao. Global land-cover classification and mapping at 30 m using quantitative remote sensing technique[D] Beijing:University of Chinese Academy of Sciences(Aerospace Information Research Institute), 2020.
30 张肖. 全球30 m地表覆盖定量遥感分类与制图研究[D].北京:中国科学院大学(中国科学院空天信息创新研究院), 2020.
31 Gong P, Wang J, Yu L, et al. Finer resolution observation and monitoring of global land cover: first mapping results with Landsat TM and ETM+data[J/OL]. International Journal of Remote Sensing,2013,34(7):2607-2654. .
32 Otsu N. Threshold selection method from gray-level histogram[J]. IEEE Transactions on Systems Man and Cybernetics,1979,9(1):62-66.
33 Breiman L. Random forests[J]. Machine Learning,2001, 45(1): 5-32. .
[1] 张飞舟,刘华亮,张立福,岑奕,孙雪剑,张红明. 高分辨率光学遥感影像变化检测算法在地震灾情调查中的应用[J]. 遥感技术与应用, 2022, 37(2): 306-318.
[2] 郭擎,朱丽娅,李安,顾铃燕. 基于NDVI变化检测的滑坡遥感精细识别[J]. 遥感技术与应用, 2022, 37(1): 17-23.
[3] 陈康明,朱旭东. 基于Google Earth Engine的南方滨海盐沼植被时空演变特征分析[J]. 遥感技术与应用, 2021, 36(4): 751-759.
[4] 李晓东,闫守刚,宋开山. 遥感监测东北地区典型湖泊湿地变化的方法研究[J]. 遥感技术与应用, 2021, 36(4): 728-741.
[5] 王崇阳,田昕. 基于GF⁃1 PMS数据的森林覆盖变化检测[J]. 遥感技术与应用, 2021, 36(1): 208-216.
[6] 罗家顺,邱建秀,赵天杰,王大刚. 基于Sentinel-1数据的黑河中游土壤水分反演[J]. 遥感技术与应用, 2020, 35(1): 23-32.
[7] 何浩,刘修国,沈永林. 基于视差的高分辨率遥感影像建筑物变化检测[J]. 遥感技术与应用, 2019, 34(6): 1315-1323.
[8] 王源,陈富龙,胡祺,唐攀攀. COSMO-SkyMed时序影像南京城市变化检测研究[J]. 遥感技术与应用, 2019, 34(5): 1054-1063.
[9] 张因果,陈芸芝. 一种基于改进土地覆盖更新方法的新增建设用地自动提取[J]. 遥感技术与应用, 2019, 34(5): 1073-1081.
[10] 申祎,王超,胡佳乐. 一种结合空间与光谱信息的改进CVA变化检测方法[J]. 遥感技术与应用, 2019, 34(4): 799-806.
[11] 冀新莹, 韦玉春, 王问尧, 方宏. 城镇区域高分辨率遥感影像地表覆盖变化检测的误差分析[J]. 遥感技术与应用, 2018, 33(5): 932-941.
[12] 姜爱辉,刘国林,陈富龙. 基于PALSAR-1影像的汉函谷关遗迹变化检测研究[J]. 遥感技术与应用, 2017, 32(5): 787-793.
[13] 张祥,陈报章,赵慧,汪磊. 基于时序Sentinel-1A数据的农田土壤水分变化检测分析[J]. 遥感技术与应用, 2017, 32(2): 338-345.
[14] 张明哲,张红,王超,刘萌,谢镭. 基于超像素分割和多方法融合的SAR 图像变化检测方法[J]. 遥感技术与应用, 2016, 31(3): 481-487.
[15] 王颖洁,刘良云,王志慧. 基于时序Landsat数据的三江平原植被地表类型变化遥感探测研究[J]. 遥感技术与应用, 2015, 30(5): 959-968.