Large margin distribution machine for hyperspectral image classification

Large margin distribution machine for hyperspectral image classification

Support vector machine (SVM) classifiers are widely applied to hyperspectral image (HSI) classification and provide significant advantages in terms of accuracy, simplicity and robustness. SVM is a well-known learning algorithm with maximizing the minimum margin. However, recent theoretical results pointed out that maximizing the minimum margin leads to a lower generalization performance than optimizing the margin distribution, and proved that the margin distribution is more important. In this paper, a large margin distribution machine (LDM) is applied to HSI classification, and optimizing the margin distribution achieves a better generalization performance than SVM. Since the raw HSI feature space is not the most effective space to representing HSI, we adopt factor analysis to learn an effective HSI feature and the learned feature are further filtered by a structure-preserved filter to fully exploit the spatial structure information of HSI. The spatial structure information is integrated in the feature learning process to obtain a better HSI feature. Then, we propose a multi-class LDM to classify the filtered HSI feature. Experimental results show that the proposed LDM with feature learning method achieves the classification performance of the state-of-the-art methods in terms of visual quality and three quantitative evaluations, and indicates that LDM has a high generalization performance.
If you use these codes, please cite the paper:
@Article{ZhanJEI2016dec,
author = {Zhan, Kun and Wang, Haibo and Huang, He and Xie, Yuange},
title = {Large margin distribution machine for hyperspectral image classification},
journal = {Journal of Electronic Imaging},
year = {2016},
volume = {25},
number = {6},
pages = {063024},
doi = {10.1117/1.JEI.25.6.063024}
}
If you have any questions, Feel free to contact with me. (Email: ice.echo#gmail.com)
http://www.escience.cn/people/kzhan

Note: Find LDM Package at: http://lamda.nju.edu.cn/code_LDM.ashx

A complete code can be find at: https://github.com/kunzhan/LDM_HSI_Classification