极化合成孔径雷达（PolSAR）以其多参数、多通道、多极化、信息记录更加完整等特点，在城市地物提取领域中发挥着重要作用，并已成为遥感影像研究领域的热点。选择覆盖苏州市的Radarsat2影像，利用极化非相干分解法和灰度共生矩阵法分别提取19种极化特征和8种纹理特征，通过分析建筑物、植被和水体的极化特征和纹理特征进行特征组合，结合主成分分析法（PCA）和支持向量机法（SVM）对城市建筑物进行提取，并定量评估精度。结果表明：基于极化特征的建筑物提取精度最高为92.4%；基于纹理特征的提取精度最高为88.9%；极化特征与纹理特征相结合可以提高精度，最高精度为93.7%；PCA特征融合算法具有较高的运算效率，同时提高了精度。
Polarimetric synthetic aperture radar (PolSAR) plays an important role in the field of building extraction because of its multi-parameter, multi-channel, multi-polarization, and rich information records. Taking Radarsat-2 image of Suzhou in 2017 as an example,19 polarization features and 8 texture features are extracted by polarization non-coherent decomposition methods and GLCM, respectively. Based on the analysis of features, we obtain the results of building extraction by PCA feature fusion and SVM algorithm. The results show that the extraction accuracies based on polarization features and texture feature are 92.4% and 88.9%, respectively. The accuracy is 93.7% when the polarization and texture features are used together. The combination of polarization and texture features improves the accuracy and the PCA feature fusion increases both efficiency and precision.
[1] Zebker H. Polarisation:applications in Remote Sensing[J]. Physics Today, 2010, 63(10):53-54.
[2] Lee J S, Pottier E. Polarimetric radar imaging:from basics to applications[M]. London:Chemical Rubber Company Press, 2009:101-175.
[3] 张腊梅, 段宝龙, 邹斌. 极化SAR图像目标分解方法的研究进展[J]. 电子与信息学报, 2016, 38(12):3 289-3 297.
[4] 陈曦, 吴涛, 陶利,等. 极化SAR发展需求及其目标识别关键技术[J]. 科技视界, 2015(16):21-22.
[5] Salehi M, Sahebi M R, Maghsoudi Y. Improving the accuracy of urban land cover classification using Radarsat-2 PolSAR data[J]. IEEE Journal of Selected Topics in Applied Earth Observations & Remote Sensing, 2014, 7(4):1 394-1 401.
[6] 顾钰培, 肖兰玲, 凌婷婷,等. 一种基于高分辨率遥感影像的建筑物提取方法[J]. 测绘与空间地理信息, 2014(4):169-171.
[7] 杨杰, 赵伶俐, 史磊,等. 基于最优极化相干系数的倾斜建筑物解译研究[J]. 测绘学报, 2012, 41(4):577-583.
[8] Lee J S, Grunes M R, Ainsworth T L, et al. Unsupervised classification using polarimetric decomposition and the complex Wishart classifier[J]. IEEE Transactions on Geoscience & Remote Sensing, 2002, 37(5):2 249-2 258.
[9] Li D, Zhang Y. Random similarity-based entropy/alpha classification of PolSAR data[J]. IEEE Journal of Selected Topics in Applied Earth Observations & Remote Sensing, 2017, 99:1-12.
[10] 何连, 秦其明, 任华忠. 一种自适应的混合Freeman/Eigenvalue极化分解模型[J]. 国土资源遥感, 2017, 29(2):8-14.
[11] Freeman A, Durden S L. A three-component scattering model for polarimetric SAR data[J]. IEEE Transactions on Geoscience & Remote Sensing, 1998, 36(3):963-973.
[12] Yamaguchi Y, Sato A, Boerner W M, et al. Four-component scattering power decomposition with rotation of coherency matrix[J]. IEEE Transactions on Geoscience & Remote Sensing, 2011, 49(6):2 251-2 258.
[13] 张腊梅. 极化SAR图像人造目标特征提取与检测方法研究[D]. 哈尔滨:哈尔滨工业大学, 2010.
[14] Zhang L, Zou B, Cai H, et al. Multiple-component scattering model for polarimetric SAR image decomposition[J]. IEEE Geoscience & Remote Sensing Letters, 2008, 5(4):603-607.
[15] Lee J S, Schuler D L, Ainsworth T L. Polarimetric SAR data compensation for terrain azimuth slope variation[J]. IEEE Transactions on Geoscience & Remote Sensing, 2000, 38(5):2 153-2 163.
[16] An W, Cui Y, Yang J, et al. Fast alternatives to H/alpha for polarimetric SAR[J]. IEEE Geoscience & Remote Sensing Letters, 2010, 7(2):343-347.
[17] 蔡永俊, 张祥坤, 姜景山. 极化SAR自适应三分量分解方法[J]. 测绘学报, 2016, 45(9):1 089-1 095.
[18] 张祥, 邓喀中, 范洪冬,等. 基于目标分解的极化SAR图像SVM监督分类[J]. 计算机应用研究, 2013, 30(1):295-298.
[19] 杜培军. RADARSAT图象滤波的研究[J]. 中国矿业大学学报, 2002, 31(2):25-30.
[20] Smith L I. A tutorial on principal components analysis[J]. Information Fusion, 2002, 51(3):52.
