欢迎访问中国科学院大学学报，今天是
RSS English

中国科学院大学学报 >

2025 , Vol. 42 >Issue 5: 686 - 699

DOI: https://doi.org/10.7523/j.ucas.2024.003

电子信息与计算机科学

基于张量表示的多时相极化SAR农作物分类方法

  • 许璐 ,
  • 张红 ,
  • 王超 ,
  • 吴樊 ,
  • 张波 ,
  • 汤益先
展开
  • 1. 中国科学院空天信息创新研究院 中国科学院数字地球重点实验室, 北京 100094;
    2. 可持续发展大数据研究中心, 北京 100094;
    3. 中国科学院大学, 北京 100049

收稿日期: 2023-11-07

  修回日期: 2024-01-17

  网络出版日期: 2024-01-25

基金资助

国家自然科学基金(42001278，41971395，41930110)资助

收起

Multitemporal polarimetric SAR crop classification method based on tensor representation

  • XU Lu ,
  • ZHANG Hong ,
  • WANG Chao ,
  • WU Fan ,
  • ZHANG Bo ,
  • TANG Yixian
Expand
  • 1. CAS Key Laboratory of Digital Earth Science, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China;
    2. The International Research Center of Big Data for Sustainable Development Goals, Beijing 100094, China;
    3. University of Chinese Academy of Sciences, Beijing 100049, China

Received date: 2023-11-07

  Revised date: 2024-01-17

  Online published: 2024-01-25

Fold

摘要

为充分利用多时相极化合成孔径雷达(SAR)数据的时间相干性和散射特征，提出一个多时相极化SAR分类方法，该方法基于完整的极化协方差矩阵，能够在张量空间保持协方差矩阵的复数矩阵结构，实现时间维度的独立表示，可同时适用于全极化和简缩极化SAR。该方法采用目标级的分类策略，首先，通过简单线性迭代聚类方法实现多时相极化SAR的超像素联合分割；随后，将目标的极化协方差矩阵表示为张量的形式，利用张量域的多线性主成分分析方法，实现多时相极化协方差矩阵的特征降维；最后，用决策树方法实现农作物分类。获取4景RADARSAT-2 Fine Quad模式全极化SAR图像，对天津市武清区农作物种植区开展作物分类实验，相较于其他文献提出的方法，本文方法取得了最高的总体分类精度。进一步，将该方法推广至π/4模式和CTLR模式的简缩极化SAR，并将其农作物分类精度与全极化SAR进行对比，以研究不同极化SAR数据对作物的识别能力。实验结果表明，简缩极化SAR可以取得与全极化SAR相当的总体分类精度，但全极化SAR在水稻、荷花等小样本地物上表现更优。

关键词： 合成孔径雷达 (SAR); 全极化 (FP); 简缩极化 (CP); 农作物分类; 张量; 多线性主成分分析 (MPCA)

本文引用格式

许璐 , 张红 , 王超 , 吴樊 , 张波 , 汤益先 . 基于张量表示的多时相极化SAR农作物分类方法[J]. 中国科学院大学学报, 2025 , 42(5) : 686 -699 . DOI: 10.7523/j.ucas.2024.003

Abstract

Multitemporal polarimetric synthetic aperture radar (SAR) provides abundant polarimetric scattering information, which is of great value to the long-term monitoring of various crop lands. To make full use of the time correlation and polarimetric information of multitemporal polarimetric SAR, this paper proposed a multitemporal polarimetric SAR crop classification method, which is based on the complete polarimetric covariance matrix. The method can maintain the complex matrix structure of covariance matrix and realize the independent representation of time dimension in tensor space, so that it can be applied to both full- and compact-polarimetric SAR. The method adopted the object-level classification strategy. Firstly, the superpixel segmentation of multitemporal SAR data was achieved by the simple linear iterative clustering (SLIC) method. Then, the covariance matrices of multitemporal SAR were expressed as tensors, and the multilinear principal component analysis (MPCA) method was used to reduce the feature dimension. Finally, the crop classification is achieved by decision tree. In this research, four multitemporal RADARSAT-2 Fine Quad SAR images covered Wuqing District，Tianjin, were used for the crop classification experiments. Compared with methods proposed in other references, the method proposed in this paper achieved the highest overall classification accuracy. Besides, the proposed method was applied to the π/4 mode and the CTLR mode compact-polarimetric SAR to discuss the capability of different kinds of polarimetric SAR in crop classification. Compared with the full-polarimetric SAR, the compact-polarimetric SAR could achieve comparable classification accuracies, but the full-polarimetric SAR performed better at the classes with small sample size, such as rice and lotus.

Key words： synthetic aperture radar (SAR); full polarimetric (FP); compact polarimetric (CP); crop classification; tensor; multilinear principal component analysis (MPCA)

参考文献

[1] El Hajj M, Baghdadi N, Bazzi H, et al. Penetration analysis of SAR signals in the C and L bands for wheat, maize, and grasslands[J]. Remote Sensing, 2018, 11(1): 31. DOI: 10.3390/rs11010031.
[2] Galvão L S, Roberts D A, Formaggio A R, et al. View angle effects on the discrimination of soybean varieties and on the relationships between vegetation indices and yield using off-nadir Hyperion data[J]. Remote Sensing of Environment, 2009, 113(4): 846-856. DOI: 10.1016/j.rse.2008.12.010.
[3] Colombo R, Meroni M, Marchesi A, et al. Estimation of leaf and canopy water content in poplar plantations by means of hyperspectral indices and inverse modeling[J]. Remote Sensing of Environment, 2008, 112(4): 1820-1834. DOI: 10.1016/j.rse.2007.09.005.
[4] Franke J, Menz G. Multi-temporal wheat disease detection by multi-spectral remote sensing[J]. Precision Agriculture, 2007, 8(3): 161-172. DOI: 10.1007/s11119-007-9036-y.
[5] Ranjitha G, Srinivasan M R, Rajesh A. Detection and estimation of damage caused by Thrips Thrips tabaci (lind) of cotton using hyperspectral radiometer[J]. Agrotechnology, 2014, 3(1): 123. DOI: 10.4172/2168-9881.1000123.
[6] Cable J W, Kovacs J M, Jiao X F, et al. Agricultural monitoring in northeastern Ontario, Canada, using multi-temporal polarimetric RADARSAT-2 data[J]. Remote Sensing, 2014, 6(3):2343-2371. DOI: 10.3390/rs6032343.
[7] Pichierri M, Hajnsek I, Zwieback S, et al. On the potential of Polarimetric SAR Interferometry to characterize the biomass, moisture and structure of agricultural crops at L-, C- and X-Bands[J]. Remote Sensing of Environment, 2018, 204: 596-616. DOI: 10.1016/j.rse.2017.09.039.
[8] Fieuzal R, Marais Sicre C, Baup F. Estimation of corn yield using multi-temporal optical and radar satellite data and artificial neural networks[J]. International Journal of Applied Earth Observation and Geoinformation, 2017, 57: 14-23. DOI: 10.1016/j.jag.2016.12.011.
[9] Larrañaga A, Álvarez -Mozos J. On the added value of quad-pol data in a multi-temporal crop classification framework based on RADARSAT-2 imagery[J]. Remote Sensing, 2016, 8(4):335. DOI: 10.3390/rs8040335.
[10] Li K, Brisco B, Yun S, et al. Polarimetric decomposition with RADARSAT-2 for rice mapping and monitoring[J]. Canadian Journal of Remote Sensing, 2012, 38(2): 169-179. DOI: 10.5589/m12-024.
[11] Jiao X F, Kovacs J M, Shang J L, et al. Object-oriented crop mapping and monitoring using multi-temporal polarimetric RADARSAT-2 data[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2014, 96: 38-46. DOI: 10.1016/j.isprsjprs.2014.06.014.
[12] McNairn H, Kross A, Lapen D, et al. Early season monitoring of corn and soybeans with TerraSAR-X and RADARSAT-2[J]. International Journal of Applied Earth Observation and Geoinformation, 2014, 28: 252-259. DOI: 10.1016/j.jag.2013.12.015.
[13] Demarez V, Helen F, Marais-Sicre C, et al. In-season mapping of irrigated crops using Landsat 8 and sentinel-1 time series[J]. Remote Sensing, 2019, 11(2): 118. DOI: 10.3390/rs11020118.
[14] Wang M, Liu C G, Han D R, et al. Assessment of GF3 full-polarimetric SAR data for dryland crop classification with different polarimetric decomposition methods[J]. Sensors, 2022, 22(16):6087. DOI: 10.3390/s22166087.
[15] Xie Q H, Dou Q, Peng X, et al. Crop classification based on the physically constrained general model-based decomposition using multi-temporal RADARSAT-2 data[J]. Remote Sensing, 2022, 14(11): 2668. DOI: 10.3390/rs14112668.
[16] Ndikumana E, Ho Tong Minh D, Baghdadi N, et al. Deep recurrent neural network for agricultural classification using multitemporal SAR Sentinel-1 for Camargue, France[J]. Remote Sensing, 2018, 10(8): 1217. DOI: 10.3390/rs10081217.
[17] Geng J, Wang H Y, Fan J C, et al. SAR image classification via deep recurrent encoding neural networks[J]. IEEE Transactions on Geoscience and Remote Sensing, 2018, 56(4): 2255-2269. DOI: 10.1109/TGRS.2017.2777868.
[18] Zhang W T, Liu L, Bai Y, et al. Crop classification based on multi-temporal PolSAR images with a single tensor network[J]. Pattern Recognition, 2023, 143: 109773. DOI: 10.1016/j.patcog.2023.109773.
[19] Yin Q, Lin Z Y, Hu W, et al. Crop classification of multitemporal PolSAR based on 3-D attention module with ViT[J]. IEEE Geoscience and Remote Sensing Letters, 2023, 20: 4005405. DOI: 10.1109/LGRS.2023.3270488.
[20] Souyris J C, Imbo P, Fjortoft R, et al. Compact polarimetry based on symmetry properties of geophysical media: the π/4 mode[J]. IEEE Transactions on Geoscience and Remote Sensing, 2005, 43(3): 634-646. DOI: 10.1109/TGRS.2004.842486.
[21] Stacy N, Preiss M. Compact polarimetric analysis of X-band SAR data[C]//6th European Conference on Synthetic Aperture Radar (EUSAR). Dresden, Germany. 2006.
[22] Raney R K. Hybrid-polarity SAR architecture[J]. IEEE Transactions on Geoscience and Remote Sensing, 2007, 45(11): 3397-3404. DOI: 10.1109/TGRS.2007.895883.
[23] Ballester-Berman J D, Lopez-Sanchez J M. Time series of hybrid-polarity parameters over agricultural crops[J]. IEEE Geoscience and Remote Sensing Letters, 2012, 9(1): 139-143. DOI: 10.1109/LGRS.2011.2162312.
[24] Shang J L, McNairn H, Charbonneau F, et al. Sensitivity analysis of compact polarimetry parameters to crop growth using simulated RADARSAT-2 SAR data[C]//2012 IEEE International Geoscience and Remote Sensing Symposium. Munich, Germany. IEEE, 2012: 1825-1828. DOI: 10.1109/IGARSS.2012.6351156.
[25] 张红, 谢镭, 王超, 等. 简缩极化SAR数据信息提取与应用[J]. 中国图象图形学报, 2013, 18(9): 1065-1073. DOI:10.11834/jig.20130902.
[26] Brisco B, Li K, Tedford B, et al. Compact polarimetry assessment for rice and wetland mapping[J]. International Journal of Remote Sensing, 2013, 34(6): 1949-1964. DOI: 10.1080/01431161.2012.730156.
[27] Xie L, Zhang H, Wu F, et al. Capability of rice mapping using hybrid polarimetric SAR data[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2015, 8(8): 3812-3822. DOI: 10.1109/JSTARS.2014.2387214.
[28] Uppala D, Venkata R K, Poloju S, et al. Discrimination of maize crop with hybrid polarimetric RISAT1 data[J]. International Journal of Remote Sensing, 2016, 37(11):2641-2652. DOI: 10.1080/01431161.2016.1184353.
[29] Dingle Robertson L, McNairn H, Jiao X F, et al. Monitoring crops using compact polarimetry and the RADARSAT constellation mission[J]. Canadian Journal of Remote Sensing, 2022, 48(6): 793-813. DOI: 10.1080/07038992.2022.2121271.
[30] Guo X Y, Yin J J, Li K, et al. Fine classification of rice paddy using multitemporal compact polarimetric SAR C band data based on machine learning methods[J]. Frontiers of Earth Science, 2023. DOI: 10.1007/s11707-022-1011-4.
[31] Ainsworth T L, Kelly J P, Lee J S. Classification comparisons between dual-pol, compact polarimetric and quad-pol SAR imagery[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2009, 64(5): 464-471. DOI: 10.1016/j.isprsjprs.2008.12.008.
[32] Guo R, He W, Zhang S X, et al. Analysis of three-component decomposition to compact polarimetric synthetic aperture radar[J]. IET Radar, Sonar & Navigation, 2014, 8(6): 685-691. DOI: 10.1049/iet-rsn.2013.0114.
[33] Zhang H, Xie L, Wang C, et al. Investigation of the capability of H-α decomposition of compact polarimetric SAR[J]. IEEE Geoscience and Remote Sensing Letters, 2014, 11(4): 868-872. DOI: 10.1109/LGRS.2013.2280456.
[34] 谢镭. 多模式极化SAR图像分解与分类方法及应用研究[D]. 北京: 中国科学院大学, 2016: 46-59.
[35] Xie L, Zhang H, Li H Z, et al. A unified framework for crop classification in Southern China using fully polarimetric, dual polarimetric, and compact polarimetric SAR data[J]. International Journal of Remote Sensing, 2015, 36(14): 3798-3818. DOI: 10.1080/01431161.2015.1070319.
[36] Lu H P, Plataniotis K N, Venetsanopoulos A N. MPCA: multilinear principal component analysis of tensor objects[J]. IEEE Transactions on Neural Networks, 2008, 19(1): 18-39. DOI: 10.1109/TNN.2007.901277.
[37] Liu B, Hu H, Wang H Y, et al. Superpixel-based classification with an adaptive number of classes for polarimetric SAR images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2013, 51(2): 907-924. DOI: 10.1109/TGRS.2012.2203358.
[38] Qin F C, Guo J M, Lang F K. Superpixel segmentation for polarimetric SAR imagery using local iterative clustering[J]. IEEE Geoscience and Remote Sensing Letters, 2015, 12(1): 13-17. DOI: 10.1109/LGRS.2014.2322960.
[39] Wang Y H, Han C Z, Tupin F. PolSAR data segmentation by combining tensor space cluster analysis and Markovian framework[J]. IEEE Geoscience and Remote Sensing Letters, 2010, 7(1): 210-214. DOI: 10.1109/LGRS.2009.2031660.
[40] Breiman L. Classification and regression trees[M]. Routledge, 1984. DOI: 10.1201/9781315139470.
[41] Liu C, Yin J J, Yang J, et al. Classification of multi-frequency polarimetric SAR images based on multi-linear subspace learning of tensor objects[J]. Remote Sensing, 2015, 7(7): 9253-9268. DOI: 10.3390/rs70709253.
[42] Tao M L, Zhou F, Liu Y, et al. Tensorial independent component analysis-based feature extraction for polarimetric SAR data classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 2015, 53(5): 2481-2495. DOI: 10.1109/TGRS.2014.2360943.
Options
文章导航

/

〈 〉

首页 | 期刊介绍 | 编委会 | 投稿指南 | 期刊浏览 | 期刊订阅 | 联系我们

版权所有 © 《中国科学院大学学报》编辑部

本系统由北京玛格泰克科技发展有限公司设计开发

技术支持：support@magtech.com.cn

访问统计

总访问量： 今日访问： 当前在线：

联系我们

通讯地址：北京玉泉路19号（甲）中国科学院大学科研楼410

联系电话:（010）88256013