Latent subclass learning-based unsupervised ensemble feature extraction method for hyperspectral image classification
Authors:
- 3 April 2015
- journal article
- research article
- Published by Taylor & Francis in Remote Sensing Letters
- Vol. 6 (4) , 257-266
Abstract
A novel unsupervised ensemble feature learning method for hyperspectral image classification is proposed in this study. Firstly, we randomly sample multiple discriminative subsets from a hyperspectral image with the novel spatially constrained similarity measurement. Each subset consists of a small amount of representative pixels. Each pixel in the subset was assigned with a latent-subclass/pseudo label. Multiple multinomial logistic regression classifiers are then adopted to build relations between pixels and their latent subclass labels, where each classifier is trained with one subset. Finally, the predicted results of different classifiers for a given pixel are assembled as its ensemble feature. More discriminative features are extracted by the proposed method compared with features extracted by traditional unsupervised methods such as principal component analysis and non-negative matrix factorization. Experimental results on hyperspectral image classification demonstrate the effectiveness of the proposed method.Keywords
Funding Information
- National Natural Science Foundation of China (61201291,61231016,61301192)
This publication has 16 references indexed in Scilit:
- Local discriminant non-negative matrix factorization feature extraction for hyperspectral image classificationInternational Journal of Remote Sensing, 2014
- A new committee-based active learning (CBAL) approach to hyperspectral remote sensing data classificationRemote Sensing Letters, 2014
- Constrained nonnegative matrix factorization and hyperspectral image dimensionality reductionRemote Sensing Letters, 2014
- Band elimination of hyperspectral imagery using partitioned band image correlation and capacitory discriminationInternational Journal of Remote Sensing, 2014
- Linear local tangent space alignment and application to face recognitionNeurocomputing, 2007
- An unsupervised artificial immune classifier for multi/hyperspectral remote sensing imageryIEEE Transactions on Geoscience and Remote Sensing, 2006
- Sparse multinomial logistic regression: fast algorithms and generalization boundsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2005
- Nonparametric weighted feature extraction for classificationIEEE Transactions on Geoscience and Remote Sensing, 2004
- Unsupervised classification of hyperspectral data: an ICA mixture model based approachInternational Journal of Remote Sensing, 2004
- Cluster-space representation for hyperspectral data classificationIEEE Transactions on Geoscience and Remote Sensing, 2002