Self-Supervised Feature Learning With CRF Embedding for Hyperspectral Image Classification
- 27 November 2018
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 57 (5) , 2628-2642
Abstract
The challenges in hyperspectral image (HSI) classification lie in the existence of noisy spectral information and lack of contextual information among pixels. Considering the three different levels in HSIs, i.e., subpixel, pixel, and superpixel, offer complementary information, we develop a novel HSI feature learning network (HSINet) to learn consistent features by self-supervision for HSI classification. HSINet contains a three-layer deep neural network and a multifeature convolutional neural network. It automatically extracts the features such as spatial, spectral, color, and boundary as well as context information. To boost the performance of self-supervised feature learning with the likelihood maximization, the conditional random field (CRF) framework is embedded into HSINet. The potential terms of unary, pairwise, and higher order in CRF are constructed by the corresponding subpixel, pixel, and superpixel. Furthermore, the feedback information derived from these terms are also fused into the different-level feature learning process, which makes the HSINet-CRF be a trainable end-to-end deep learning model with the back-propagation algorithm. Comprehensive evaluations are performed on three widely used HSI data sets and our method outperforms the state-of-the-art methods.Keywords
Funding Information
- National Basic Research Program of China (2018YFC0213600)
- National Natural Science Foundation of China (41371324, 41801241)
- China Postdoctoral Science Foundation (2016M600953, 2018T110065)
This publication has 46 references indexed in Scilit:
- Efficient Superpixel-Level Multitask Joint Sparse Representation for Hyperspectral Image ClassificationIEEE Transactions on Geoscience and Remote Sensing, 2015
- MRM-Lasso: A Sparse Multiview Feature Selection Method via Low-Rank AnalysisIEEE Transactions on Neural Networks and Learning Systems, 2015
- Spectral–Spatial Classification of Hyperspectral Images With a Superpixel-Based Discriminative Sparse ModelIEEE Transactions on Geoscience and Remote Sensing, 2015
- Complementarity of Discriminative Classifiers and Spectral Unmixing Techniques for the Interpretation of Hyperspectral ImagesIEEE Transactions on Geoscience and Remote Sensing, 2014
- SEEDS: Superpixels Extracted Via Energy-Driven SamplingInternational Journal of Computer Vision, 2014
- Hyperspectral Image Classification Through Bilayer Graph-Based LearningIEEE Transactions on Image Processing, 2014
- Spectral–Spatial Classification of Hyperspectral Images Based on Hidden Markov Random FieldsIEEE Transactions on Geoscience and Remote Sensing, 2013
- Local Manifold Learning-Based $k$ -Nearest-Neighbor for Hyperspectral Image ClassificationIEEE Transactions on Geoscience and Remote Sensing, 2010
- Segmentation and classification of hyperspectral images using watershed transformationPattern Recognition, 2010
- Normalized cuts and image segmentationIEEE Transactions on Pattern Analysis and Machine Intelligence, 2000