Learning Rotation Domain Deep Mutual Information Using Convolutional LSTM for Unsupervised PolSAR Image Classification
Open Access
- 12 December 2020
- journal article
- research article
- Published by MDPI AG in Remote Sensing
- Vol. 12 (24) , 4075
Abstract
Deep learning can archive state-of-the-art performance in polarimetric synthetic aperture radar (PolSAR) image classification with plenty of labeled data. However, obtaining large number of accurately labeled samples of PolSAR data is very hard, which limits the practical use of deep learning. Therefore, unsupervised PolSAR image classification is worthy of further investigation that is based on deep learning. Inspired by the superior performance of deep mutual information in natural image feature learning and clustering, an end-to-end Convolutional Long Short Term Memory (ConvLSTM) network is used in order to learn the deep mutual information of polarimetric coherent matrices in the rotation domain with different polarimetric orientation angles (POAs) for unsupervised PolSAR image classification. First, for each pixel, paired “POA-spatio” samples are generated from the polarimetric coherent matrices with different POAs. Second, a special designed ConvLSTM network, along with deep mutual information losses, is used in order to learn the discriminative deep mutual information feature representation of the paired data. Finally, the classification results can be output directly from the trained network model. The proposed method is trained in an end-to-end manner and does not have cumbersome pipelines. Experiments on four real PolSAR datasets show that the performance of proposed method surpasses some state-of-the-art deep learning unsupervised classification methods.Keywords
Funding Information
- National Natural Science Foundation of China (62071336)
- National Key Research and Development Program of China (2018YFB2100503)
- Chinese Technology Research and Development of the Major Project of High-Resolution Earth Observation System (03-Y20A10-9001-15/16)
- Thirteen-Five Civil Aerospace Planning Project — Integration of Communication, Navigation and Remote Sensing Comprehensive Application Technology
This publication has 41 references indexed in Scilit:
- Adaptive unsupervised classification of polarimetric SAR images using the improved affinity propagation clustering algorithmInternational Journal of Remote Sensing, 2016
- Unsupervised Classification of PolSAR Imagery via Kernel Sparse Subspace ClusteringIEEE Geoscience and Remote Sensing Letters, 2016
- Random Forest and Rotation Forest for fully polarized SAR image classification using polarimetric and spatial featuresISPRS Journal of Photogrammetry and Remote Sensing, 2015
- Unsupervised classification of polarimetric SAR imagery using large-scale spectral clustering with spatial constraintsInternational Journal of Remote Sensing, 2015
- Deep learning in neural networks: An overviewNeural Networks, 2015
- Land Cover Classification for Polarimetric SAR Images Based on Mixture ModelsRemote Sensing, 2014
- Uniform Polarimetric Matrix Rotation Theory and Its ApplicationsIEEE Transactions on Geoscience and Remote Sensing, 2013
- Fusion of Support Vector Machines for Classification of Multisensor DataIEEE Transactions on Geoscience and Remote Sensing, 2007
- An Unsupervised Segmentation With an Adaptive Number of Clusters Using the $SPAN/H/\alpha/A$ Space and the Complex Wishart Clustering for Fully Polarimetric SAR Data AnalysisIEEE Transactions on Geoscience and Remote Sensing, 2007
- Quantitative comparison of classification capability: fully polarimetric versus dual and single-polarization SARIEEE Transactions on Geoscience and Remote Sensing, 2001