Recursive Autoencoders-Based Unsupervised Feature Learning for Hyperspectral Image Classification

For hyperspectral image (HSI) classification, it is very important to learn effective features for the discrimination purpose. Meanwhile, the ability to combine spectral and spatial information together in a deep level is also important for feature learning. In this letter, we propose an unsupervised feature learning method for HSI classification, which is based on recursive autoencoders (RAE) network. RAE utilizes the spatial and spectral information and produces high-level features from the original data. It learns features from the neighborhood of the investigated pixel to represent the whole local homogeneous area of the image. In addition, to obtain more accurate representation of the investigated pixel, a weighting scheme is adopted based on the neighboring pixels, where the weights are determined by the spectral similarity between the neighboring pixels and the investigated pixel. The effectiveness of our method is evaluated by the experiments on two hyperspectral data sets, and the results show that our proposed method has a better performance.

[1]  Carlo Gatta,et al.  Unsupervised Deep Feature Extraction for Remote Sensing Image Classification , 2015, IEEE Transactions on Geoscience and Remote Sensing.

[2]  Andrew Y. Ng,et al.  Parsing Natural Scenes and Natural Language with Recursive Neural Networks , 2011, ICML.

[3]  Gang Wang,et al.  Deep Learning-Based Classification of Hyperspectral Data , 2014, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[4]  FauvelMathieu,et al.  Kernel principal component analysis for the classification of hyperspectral remote sensing data over urban areas , 2009 .

[5]  Lorenzo Bruzzone,et al.  Classification of hyperspectral remote sensing images with support vector machines , 2004, IEEE Transactions on Geoscience and Remote Sensing.

[6]  Qi Wang,et al.  Salient Band Selection for Hyperspectral Image Classification via Manifold Ranking , 2016, IEEE Transactions on Neural Networks and Learning Systems.

[7]  Jeffrey Pennington,et al.  Semi-Supervised Recursive Autoencoders for Predicting Sentiment Distributions , 2011, EMNLP.

[8]  Biao Hou,et al.  Spatial-spectral classification based on group sparse coding for hyperspectral image , 2013, 2013 IEEE International Geoscience and Remote Sensing Symposium - IGARSS.

[9]  Jon Atli Benediktsson,et al.  Spectral–Spatial Classification of Hyperspectral Imagery Based on Partitional Clustering Techniques , 2009, IEEE Transactions on Geoscience and Remote Sensing.

[10]  Jorge Nocedal,et al.  On the limited memory BFGS method for large scale optimization , 1989, Math. Program..

[11]  Yangyang Li,et al.  PSO-based automatic relevance determination and feature selection system for hyperspectral image classification , 2012 .

[12]  Dimitris G. Manolakis,et al.  Detection algorithms for hyperspectral imaging applications , 2002, IEEE Signal Process. Mag..

[13]  Nasser M. Nasrabadi,et al.  Automated Hyperspectral Cueing for Civilian Search and Rescue , 2009, Proceedings of the IEEE.

[14]  Tim R. McVicar,et al.  Preprocessing EO-1 Hyperion hyperspectral data to support the application of agricultural indexes , 2003, IEEE Trans. Geosci. Remote. Sens..

[15]  Yee Whye Teh,et al.  A Fast Learning Algorithm for Deep Belief Nets , 2006, Neural Computation.

[16]  Xiangrong Zhang,et al.  Semisupervised Dimensionality Reduction of Hyperspectral Images via Local Scaling Cut Criterion , 2013, IEEE Geoscience and Remote Sensing Letters.

[17]  Qi Wang,et al.  Hyperspectral Image Classification via Multitask Joint Sparse Representation and Stepwise MRF Optimization , 2016, IEEE Transactions on Cybernetics.

[18]  Glenn Healey,et al.  Hyperspectral Region Classification Using a Three-Dimensional Gabor Filterbank , 2010, IEEE Transactions on Geoscience and Remote Sensing.

[19]  Kunihiko Fukushima,et al.  Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position , 1980, Biological Cybernetics.

[20]  Fuan Tsai,et al.  Feature Extraction of Hyperspectral Image Cubes Using Three-Dimensional Gray-Level Cooccurrence , 2013, IEEE Transactions on Geoscience and Remote Sensing.

[21]  Jon Atli Benediktsson,et al.  Kernel Principal Component Analysis for the Classification of Hyperspectral Remote Sensing Data over Urban Areas , 2009, EURASIP J. Adv. Signal Process..

[22]  Trac D. Tran,et al.  Exploiting Sparsity in Hyperspectral Image Classification via Graphical Models , 2013, IEEE Geoscience and Remote Sensing Letters.

[23]  J. Shan,et al.  Principal Component Analysis for Hyperspectral Image Classification , 2002 .

[24]  Yoshua Bengio,et al.  Greedy Layer-Wise Training of Deep Networks , 2006, NIPS.

[25]  Xiuping Jia,et al.  Deep Feature Extraction and Classification of Hyperspectral Images Based on Convolutional Neural Networks , 2016, IEEE Transactions on Geoscience and Remote Sensing.

[26]  Jon Atli Benediktsson,et al.  Generalized Composite Kernel Framework for Hyperspectral Image Classification , 2013, IEEE Transactions on Geoscience and Remote Sensing.

[27]  Johannes R. Sveinsson,et al.  Spectral and spatial classification of hyperspectral data using SVMs and morphological profiles , 2008, 2007 IEEE International Geoscience and Remote Sensing Symposium.

[28]  Trac D. Tran,et al.  Hyperspectral Image Classification Using Dictionary-Based Sparse Representation , 2011, IEEE Transactions on Geoscience and Remote Sensing.