谢凯浪;赵泉华;李玉;

• 1:辽宁工程技术大学测绘与地理科学学院遥感科学与应用研究所

摘要(Abstract)：

针对极化合成孔径雷达(PolSAR)影像地物分类特征表征性弱,以及传统全卷积网络(FCN)分类精度较低、效果差的问题,该文提出了一种结合FCN和多特征的全极化SAR土地覆盖分类算法。首先,根据PolSAR影像和极化目标分解获取散射特征参数构建特征空间,利用主成分分析(PCA)对特征空间实现降维,以优化特征组合;接着,以SegNet建模思想为基础,在网络中层嵌入多层多尺度非对称卷积单元(MACU)结构,并在中层添加代价函数构建双代价收敛(DC)结构,基于此设计了DC-MA-FCN网络;然后,以优化后的特征组合为输入,通过DC-MA-FCN网络进行多层自主学习训练网络,并利用训练好的网络进行PolSAR影像初始分类;最后,组合DC-MA-FCN网络分类结果和形态学方法实现最终分类。该方法对两地区的PolSAR影像进行取样和试验,并使用多种评价指标定量分析,表明了算法的可行性和有效性。

关键词(KeyWords)： 极化SAR;全卷积网络;多尺度非对称卷积单元;代价函数

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金青年科学基金项目(41301479);; 辽宁省高校创新人才支持计划项目(LR2016061);; 辽宁省教育厅科学技术研究一般项目(LJCL009)

作者(Authors): 谢凯浪;赵泉华;李玉;

DOI: 10.16251/j.cnki.1009-2307.2020.01.012

参考文献(References)：

• [1]张庆君．高分三号卫星总体设计与关键技术[J]．测绘学报，2017,46(3):269-277.(ZHANG Qingjun.System design and key technologies of the GF-3satellite[J].Acta Geodaetica et Cartographica Sinica,2017,46(3):269-277.)
• [2] CHEN Y Y,HE X F,WANG J,et al.The influence of polarimetric parameters and an object-based approach on land cover classification in coastal wetlands[J].Remote Sensing,2014,6(12):12575-12592.
• [3]汪洋，鲁加国，张长耀．基于Krogager分解和SVM的极化SAR图像分类[J]．遥感技术与应用，2007,22(1):70-74.(WANG Yang,LU Jiaguo,ZHANG Changyao.Classification of polarimetric SAR image based on Krogager decomposition and SVM[J].Remote Sensing Technology and Application,2007,22(1):70-74.)
• [4]郎丰铠，杨杰，赵伶俐，等．基于Freeman散射熵和各向异性度的极化SAR影像分类算法研究[J]．测绘学报，2012,41(4):556-562.(LANG Fengkai,YANG Jie,ZHAO Lingli,et al.Polarimetric SAR data classification with Freeman entropy and anisotropy analysis[J].Acta Geodaetica et Cartographica Sinica,2012,41(4):556-562.)
• [5] LEE J S,GRUNES M R.Classification of multi-look polarimetric SAR data based on complex wishart distribution[C]//National Telesystems Conference.Piscataway:IEEE,1992:2299-2311.
• [6] LI H T,GU H Y,HAN Y S,et al.Object-oriented classification of polarimetric SAR imagery based on statistical region merging and support vector machine[C]//2008International Workshop on Earth Observation and Remote Sensing Application.Piscataway:IEEE,2008.
• [7] ZHOU Y,WANG H P,XU F,et al.Polarimetric SAR image classification using deep convolutional neural networks[J].IEEE Geoscience and Remote Sensing Letters,2017,13(12):1935-1939.
• [8] LONG J,SHELHAMER E,DARRELL T.Fully convolutional networks for semantic segmentation[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2014,39(4):640-651.
• [9] BADRINARAYANAN V,KENDALL A,CIPOLLA R.SegNet:a deep convolutional Encoder-Decoder architecture for scene segmentation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(12):2481-2495.
• [10]MIKA S,SCHOLKOPF B,SMOLA A,et al.Kernel PCA and de-noising in feature spaces[C]//Proceedings of the 1998Conference on Advances in Neural Information Processing Systems II.Cambridge:MIT Press,1998:536-542.
• [11]SZEGEDY C,VANHOUCKE V,IOFFE S,et al.Rethinking the inception architecture for computer vision[J/OL].[2019-10-30].https：//arxiv.org/pdf/1512.00567.pdf.
• [12]KINGMA D P,BA J.Adam:a method for stochastic optimization[C/OL]//International Conference on Learning Representations,2015[2019-10-30].https：//arxiv.org/pdf/1412.6980.pdf.
• [13]IOFFE S,SZEGEDY C.Batch normalization:accelerating deep network training by reducing internal covariate shift[C/OL].[2019-10-30].https：//arxiv.org/pdf/1502.03167.pdf.
• [14]SU L Z,GONG M G,SUN B,et al.Unsupervised change detection in SAR images based on locally fitting model and semi-EM algorithm[J].International Journal of Remote Sensing,2014,35(2):621-650.