Semi-Supervised Classification for Hyperspectral Images Based on Multiple Classifiers and Relaxation Strategy
Hyperspectral image (HSI) classification is a fundamental and challenging problem in remote sensing and its various applications. However, it is difficult to perfectly classify remotely sensed hyperspectral data by directly using classification techniques developed in pattern recognition. This is pa...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2018-07-01
|
Series: | ISPRS International Journal of Geo-Information |
Subjects: | |
Online Access: | http://www.mdpi.com/2220-9964/7/7/284 |
id |
doaj-bedd783bc024487997a089eab245f039 |
---|---|
record_format |
Article |
spelling |
doaj-bedd783bc024487997a089eab245f0392020-11-25T00:08:19ZengMDPI AGISPRS International Journal of Geo-Information2220-99642018-07-017728410.3390/ijgi7070284ijgi7070284Semi-Supervised Classification for Hyperspectral Images Based on Multiple Classifiers and Relaxation StrategyFuding Xie0Dongcui Hu1Fangfei Li2Jun Yang3Deshan Liu4College of Urban and Environment, Liaoning Normal University, Dalian 116029, ChinaCollege of Urban and Environment, Liaoning Normal University, Dalian 116029, ChinaCollege of Urban and Environment, Liaoning Normal University, Dalian 116029, ChinaCollege of Urban and Environment, Liaoning Normal University, Dalian 116029, ChinaCollege of Computer Science, Liaoning Normal University, Dalian 116081, ChinaHyperspectral image (HSI) classification is a fundamental and challenging problem in remote sensing and its various applications. However, it is difficult to perfectly classify remotely sensed hyperspectral data by directly using classification techniques developed in pattern recognition. This is partially owing to a multitude of noise points and the limited training samples. Based on multinomial logistic regression (MLR), the local mean-based pseudo nearest neighbor (LMPNN) rule, and the discontinuity preserving relaxation (DPR) method, in this paper, a semi-supervised method for HSI classification is proposed. In pre-processing and post-processing, the DPR strategy is adopted to denoise the original hyperspectral data and improve the classification accuracy, respectively. The application of two classifiers, MLR and LMPNN, can automatically acquire more labeled samples in terms of a few labeled instances per class. This is termed the pre-classification procedure. The final classification result of the HSI is obtained by employing the MLRsub approach. The effectiveness of the proposal is experimentally evaluated by two real hyperspectral datasets, which are widely used to test the performance of the HSI classification algorithm. The comparison results using several competing methods confirm that the proposed method is effective, even for limited training samples.http://www.mdpi.com/2220-9964/7/7/284hyperspectral imagesemi-supervised classificationmultinomial logistic regressionlocal mean-based pseudo nearest neighbordiscontinuity preserving relaxation |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Fuding Xie Dongcui Hu Fangfei Li Jun Yang Deshan Liu |
spellingShingle |
Fuding Xie Dongcui Hu Fangfei Li Jun Yang Deshan Liu Semi-Supervised Classification for Hyperspectral Images Based on Multiple Classifiers and Relaxation Strategy ISPRS International Journal of Geo-Information hyperspectral image semi-supervised classification multinomial logistic regression local mean-based pseudo nearest neighbor discontinuity preserving relaxation |
author_facet |
Fuding Xie Dongcui Hu Fangfei Li Jun Yang Deshan Liu |
author_sort |
Fuding Xie |
title |
Semi-Supervised Classification for Hyperspectral Images Based on Multiple Classifiers and Relaxation Strategy |
title_short |
Semi-Supervised Classification for Hyperspectral Images Based on Multiple Classifiers and Relaxation Strategy |
title_full |
Semi-Supervised Classification for Hyperspectral Images Based on Multiple Classifiers and Relaxation Strategy |
title_fullStr |
Semi-Supervised Classification for Hyperspectral Images Based on Multiple Classifiers and Relaxation Strategy |
title_full_unstemmed |
Semi-Supervised Classification for Hyperspectral Images Based on Multiple Classifiers and Relaxation Strategy |
title_sort |
semi-supervised classification for hyperspectral images based on multiple classifiers and relaxation strategy |
publisher |
MDPI AG |
series |
ISPRS International Journal of Geo-Information |
issn |
2220-9964 |
publishDate |
2018-07-01 |
description |
Hyperspectral image (HSI) classification is a fundamental and challenging problem in remote sensing and its various applications. However, it is difficult to perfectly classify remotely sensed hyperspectral data by directly using classification techniques developed in pattern recognition. This is partially owing to a multitude of noise points and the limited training samples. Based on multinomial logistic regression (MLR), the local mean-based pseudo nearest neighbor (LMPNN) rule, and the discontinuity preserving relaxation (DPR) method, in this paper, a semi-supervised method for HSI classification is proposed. In pre-processing and post-processing, the DPR strategy is adopted to denoise the original hyperspectral data and improve the classification accuracy, respectively. The application of two classifiers, MLR and LMPNN, can automatically acquire more labeled samples in terms of a few labeled instances per class. This is termed the pre-classification procedure. The final classification result of the HSI is obtained by employing the MLRsub approach. The effectiveness of the proposal is experimentally evaluated by two real hyperspectral datasets, which are widely used to test the performance of the HSI classification algorithm. The comparison results using several competing methods confirm that the proposed method is effective, even for limited training samples. |
topic |
hyperspectral image semi-supervised classification multinomial logistic regression local mean-based pseudo nearest neighbor discontinuity preserving relaxation |
url |
http://www.mdpi.com/2220-9964/7/7/284 |
work_keys_str_mv |
AT fudingxie semisupervisedclassificationforhyperspectralimagesbasedonmultipleclassifiersandrelaxationstrategy AT dongcuihu semisupervisedclassificationforhyperspectralimagesbasedonmultipleclassifiersandrelaxationstrategy AT fangfeili semisupervisedclassificationforhyperspectralimagesbasedonmultipleclassifiersandrelaxationstrategy AT junyang semisupervisedclassificationforhyperspectralimagesbasedonmultipleclassifiersandrelaxationstrategy AT deshanliu semisupervisedclassificationforhyperspectralimagesbasedonmultipleclassifiersandrelaxationstrategy |
_version_ |
1725415748651712512 |