[1]李林,田馨,翁永玲.基于极化SAR和光学影像特征的土地覆盖分类[J].东南大学学报(自然科学版),2021,51(3):529-534.[doi:10.3969/j.issn.1001-0505.2021.03.023]
 Li Lin,Tian Xin,Weng Yongling.Land cover classification based on polarization SAR and optical image features[J].Journal of Southeast University (Natural Science Edition),2021,51(3):529-534.[doi:10.3969/j.issn.1001-0505.2021.03.023]
点击复制

基于极化SAR和光学影像特征的土地覆盖分类()
分享到:

《东南大学学报(自然科学版)》[ISSN:1001-0505/CN:32-1178/N]

卷:
51
期数:
2021年第3期
页码:
529-534
栏目:
测绘与导航
出版日期:
2021-05-20

文章信息/Info

Title:
Land cover classification based on polarization SAR and optical image features
作者:
李林田馨翁永玲
东南大学交通学院, 南京 211189
Author(s):
Li Lin Tian Xin Weng Yongling
School of Transportation, Southeast University, Nanjing 211189, China
关键词:
合成孔径雷达 多光谱 土地覆盖分类 随机森林 特征变量
Keywords:
synthetic aperture radar(SAR) multispectral land cover classification random forest characteristic variable
分类号:
P237
DOI:
10.3969/j.issn.1001-0505.2021.03.023
摘要:
为了更好地利用多源遥感影像参与土地覆盖分类,采用一种基于合成孔径雷达(SAR)影像和光学影像相结合的特征分析及筛选方法.该方法在对光学影像和SAR影像的各类特征变量进行可分性分析后,使用随机森林算法对其组成的高维变量空间进行降维筛选,将筛选出的特征变量用于土地覆盖分类,并对实验结果进行分析比较.实验结果表明:利用随机森林算法对特征变量进行分析筛选后的变量组合可以取得最优的分类结果,总体精度和Kappa系数可以达到92.1%和0.91,相比于仅用SAR影像特征变量进行分类时分别提升了11.9%和16.7%.该方法能够充分发挥光学影像和SAR影像各自的优势,提高特征变量的利用率,使分类结果更加稳定和精确.
Abstract:
In order to make better use of multi-source remote sensing images to participate in land cover classification, a feature analysis and screening method based on the combination of synthetic aperture radar(SAR)images and optical images is adopted. After analyzing the separability of various feature variables of the optical image and SAR image, the random forest algorithm is used to reduce the dimensionality of its high-dimensional variable space. The selected feature variables are used in land cover classification, and the experimental results are analyzed and compared. Experimental results show that the variable combination after the analysis and screening of the characteristic variables using the random forest algorithm can obtain the best classification results. The overall accuracy and Kappa coefficient can reach 92.1% and 0.91. Compared with the classification using only the SAR image characteristic variables, they increase by 11.9% and 16.7%, respectively. The proposed method can give full play to the respective advantages of optical and SAR images, improves the utilization of feature variables, and makes the classification results more stable and accurate.

参考文献/References:

[1] 陈军,陈利军,李然,等.基于GlobeLand30的全球城乡建设用地空间分布与变化统计分析[J].测绘学报,2015,44(11):1181-1188. DOI:10.11947/j.AGCS.2015.20140677.
Chen J, Chen L J, Li R, et al. Spatial distribution and ten years change of global built-up areas derived from GlobeLand30[J].Acta Geodaetica et Cartographica Sinica, 2015, 44(11): 1181-1188. DOI:10.11947/j.AGCS.2015.20140677. (in Chinese)
[2] di Vittorio C A, Georgakakos A P. Land cover classification and wetland inundation mapping using MODIS[J].Remote Sensing of Environment, 2018, 204: 1-17. DOI:10.1016/j.rse.2017.11.001.
[3] Rossi C, Erten E. Paddy-rice monitoring using TanDEM-X[J].IEEE Transactions on Geoscience and Remote Sensing, 2015, 53(2): 900-910. DOI:10.1109/TGRS.2014.2330377.
[4] Erinjery J J, Singh M, Kent R. Mapping and assessment of vegetation types in the tropical rainforests of the Western Ghats using multispectral Sentinel-2 and SAR Sentinel-1 satellite imagery[J].Remote Sensing of Environment, 2018, 216: 345-354. DOI:10.1016/j.rse.2018.07.006.
[5] Shuai G Y, Zhang J S, Basso B, et al. Multi-temporal RADARSAT-2 polarimetric SAR for maize mapping supported by segmentations from high-resolution optical image[J].International Journal of Applied Earth Observation and Geoinformation, 2019, 74: 1-15. DOI:10.1016/j.jag.2018.08.021.
[6] 赵诣,蒋弥.极化SAR参数优化与光学波谱相结合的面向对象土地覆盖分类[J].测绘学报,2019,48(5):609-617. DOI:10.11947/j.AGCS.2019.20170746.
Zhao Y, Jiang M. Integration of SAR polarimetric parameters and multi-spectral data for object-based land cover classification[J].Acta Geodaetica et Cartographica Sinica, 2019, 48(5): 609-617. DOI:10.11947/j.AGCS.2019.20170746. (in Chinese)
[7] Gamba P, Aldrighi M. SAR data classification of urban areas by means of segmentation techniques and ancillary optical data[J].IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2012, 5(4): 1140-1148. DOI:10.1109/JSTARS.2012.2195774.
[8] Zhang Y Z, Zhang H S, Lin H. Improving the impervious surface estimation with combined use of optical and SAR remote sensing images[J].Remote Sensing of Environment, 2014, 141: 155-167. DOI:10.1016/j.rse.2013.10.028.
[9] Zhang H S, Xu R. Exploring the optimal integration levels between SAR and optical data for better urban land cover mapping in the Pearl River Delta[J].International Journal of Applied Earth Observation and Geoinformation, 2018, 64: 87-95. DOI:10.1016/j.jag.2017.08.013.
[10] 杨宁.高分辨率影像面向对象分类特征选择方法研究[D].西安:西安科技大学,2012.
  Yang N. Feature selection for object-oriented classification of high resolution remote sensing images[D]. Xi’an: Xi’an University of Science and Technology, 2012.(in Chinese)
[11] Luo J Q, Xu T F, Pan T, et al. An efficient method of hyperspectral image dimension reduction based on low rank representation and locally linear embedding[J].Integrated Ferroelectrics, 2020, 208(1): 206-214. DOI:10.1080/10584587.2020.1728626.
[12] 杨帆,余旭初,杨其淼,等.不同降维策略下的高光谱影像多特征分类[J].测绘与空间地理信息,2021,44(1):38-42.
  Yang F, Yu X C, Yang Q M, et al. Multi-feature classification of hyperspectral images with different dimensionality reduction strategies[J].Geomatics & Spatial Information Technology, 2021, 44(1): 38-42.(in Chinese)
[13] 杨珺雯,张锦水,朱秀芳,等.随机森林在高光谱遥感数据中降维与分类的应用[J].北京师范大学学报(自然科学版),2015,51(z1):82-88. DOI:10.16360/j.cnki.jbnuns.2015.s1.013.
Yang J W, Zhang J S, Zhu X F, et al. Random forest applied for dimension reduction and classification in hyperspectral data[J].Journal of Beijing Normal University(Natural Science), 2015, 51(z1): 82-88. DOI:10.16360/j.cnki.jbnuns.2015.s1.013. (in Chinese)
[14] 翁永玲,田庆久,惠凤鸣.IKONOS高分辨率遥感影像自身融合效果分析[J].东南大学学报(自然科学版),2004,34(2):274-277. DOI: 10.3321/j.issn:1001-0505.2004.02.031.
Weng Y L, Tian Q J, Hui F M. Research on the fusion effects of IKONOS high-resolution remote sensing image[J].Journal of Southeast University(Natural Science Edition), 2004, 34(2): 274-277. DOI:10.3321/j.issn:1001-0505.2004.02.031. (in Chinese)
[15] Lee J S, Grunes M R, Pottier E, et al. Unsupervised terrain classification preserving polarimetric scattering characteristics[J].IEEE Transactions on Geoscience and Remote Sensing, 2004, 42(4): 722-731. DOI:10.1109/TGRS.2003.819883.
[16] 田馨,廖明生.精确提取InSAR时间去相关分量的方法[J].红外与毫米波学报,2016,35(4):454-461. DOI: 10.11972/j.issn.1001-9014.2016.04.013.
Tian X, Liao M S. Accurate extraction of InSAR temporal decorrelation component[J].Journal of Infrared and Millimeter Waves, 2016, 35(4): 454-461. DOI:10.11972/j.issn.1001-9014.2016.04.013. (in Chinese)
[17] Joshi N, Baumann M, Ehammer A, et al. A review of the application of optical and radar remote sensing data fusion to land use mapping and monitoring[J].Remote Sensing, 2016, 8(1): 70. DOI:10.3390/rs8010070.

备注/Memo

备注/Memo:
收稿日期: 2020-11-03.
作者简介: 李林(1996—),女,硕士生;田馨(联系人),女,博士,副教授,tianxin@seu.edu.cn.
基金项目: 国家自然科学基金资助项目(41801244,41471352).
引用本文: 李林,田馨,翁永玲.基于极化SAR和光学影像特征的土地覆盖分类[J].东南大学学报(自然科学版),2021,51(3):529-534. DOI:10.3969/j.issn.1001-0505.2021.03.023.
更新日期/Last Update: 2021-05-20