The edge-preservation multi-classifier relearning framework for the classification of high-resolution remotely sensed imagery

Xiaopeng Han, Xin Huang (Corresponding Author), Jiayi Li, Yansheng Li, Michael Ying Yang, Jianya Gong

Research output: Contribution to journalArticleAcademicpeer-review

4 Citations (Scopus)

Abstract

In recent years, the availability of high-resolution imagery has enabled more detailed observation of the Earth. However, it is imperative to simultaneously achieve accurate interpretation and preserve the spatial details for the classification of such high-resolution data. To this aim, we propose the edge-preservation multi-classifier relearning framework (EMRF). This multi-classifier framework is made up of support vector machine (SVM), random forest (RF), and sparse multinomial logistic regression via variable splitting and augmented Lagrangian (LORSAL) classifiers, considering their complementary characteristics. To better characterize complex scenes of remote sensing images, relearning based on landscape metrics is proposed, which iteratively quantizes both the landscape composition and spatial configuration by the use of the initial classification results. In addition, a novel tri-training strategy is proposed to solve the over-smoothing effect of relearning by means of automatic selection of training samples with low classification certainties, which always distribute in or near the edge areas. Finally, EMRF flexibly combines the strengths of relearning and tri-training via the classification certainties calculated by the probabilistic output of the respective classifiers. It should be noted that, in order to achieve an unbiased evaluation, we assessed the classification accuracy of the proposed framework using both edge and non-edge test samples. The experimental results obtained with four multispectral high-resolution images confirm the efficacy of the proposed framework, in terms of both edge and non-edge accuracy.

Original languageEnglish
Pages (from-to)57-73
Number of pages17
JournalISPRS journal of photogrammetry and remote sensing
Volume138
Early online date15 Feb 2018
DOIs
Publication statusPublished - 1 Apr 2018

Fingerprint

classifiers
imagery
Classifiers
high resolution
education
logistics
image resolution
Image resolution
smoothing
availability
Support vector machines
Logistics
remote sensing
regression analysis
Remote sensing
Earth (planet)
Availability
evaluation
output
configurations

Keywords

  • ITC-ISI-JOURNAL-ARTICLE

Cite this

@article{5c38b7c0aae7460593c6a37ba2715c28,
title = "The edge-preservation multi-classifier relearning framework for the classification of high-resolution remotely sensed imagery",
abstract = "In recent years, the availability of high-resolution imagery has enabled more detailed observation of the Earth. However, it is imperative to simultaneously achieve accurate interpretation and preserve the spatial details for the classification of such high-resolution data. To this aim, we propose the edge-preservation multi-classifier relearning framework (EMRF). This multi-classifier framework is made up of support vector machine (SVM), random forest (RF), and sparse multinomial logistic regression via variable splitting and augmented Lagrangian (LORSAL) classifiers, considering their complementary characteristics. To better characterize complex scenes of remote sensing images, relearning based on landscape metrics is proposed, which iteratively quantizes both the landscape composition and spatial configuration by the use of the initial classification results. In addition, a novel tri-training strategy is proposed to solve the over-smoothing effect of relearning by means of automatic selection of training samples with low classification certainties, which always distribute in or near the edge areas. Finally, EMRF flexibly combines the strengths of relearning and tri-training via the classification certainties calculated by the probabilistic output of the respective classifiers. It should be noted that, in order to achieve an unbiased evaluation, we assessed the classification accuracy of the proposed framework using both edge and non-edge test samples. The experimental results obtained with four multispectral high-resolution images confirm the efficacy of the proposed framework, in terms of both edge and non-edge accuracy.",
keywords = "ITC-ISI-JOURNAL-ARTICLE",
author = "Xiaopeng Han and Xin Huang and Jiayi Li and Yansheng Li and Yang, {Michael Ying} and Jianya Gong",
year = "2018",
month = "4",
day = "1",
doi = "10.1016/j.isprsjprs.2018.02.009",
language = "English",
volume = "138",
pages = "57--73",
journal = "ISPRS journal of photogrammetry and remote sensing",
issn = "0924-2716",
publisher = "Elsevier",

}

The edge-preservation multi-classifier relearning framework for the classification of high-resolution remotely sensed imagery. / Han, Xiaopeng; Huang, Xin (Corresponding Author); Li, Jiayi; Li, Yansheng; Yang, Michael Ying; Gong, Jianya.

In: ISPRS journal of photogrammetry and remote sensing, Vol. 138, 01.04.2018, p. 57-73.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - The edge-preservation multi-classifier relearning framework for the classification of high-resolution remotely sensed imagery

AU - Han, Xiaopeng

AU - Huang, Xin

AU - Li, Jiayi

AU - Li, Yansheng

AU - Yang, Michael Ying

AU - Gong, Jianya

PY - 2018/4/1

Y1 - 2018/4/1

N2 - In recent years, the availability of high-resolution imagery has enabled more detailed observation of the Earth. However, it is imperative to simultaneously achieve accurate interpretation and preserve the spatial details for the classification of such high-resolution data. To this aim, we propose the edge-preservation multi-classifier relearning framework (EMRF). This multi-classifier framework is made up of support vector machine (SVM), random forest (RF), and sparse multinomial logistic regression via variable splitting and augmented Lagrangian (LORSAL) classifiers, considering their complementary characteristics. To better characterize complex scenes of remote sensing images, relearning based on landscape metrics is proposed, which iteratively quantizes both the landscape composition and spatial configuration by the use of the initial classification results. In addition, a novel tri-training strategy is proposed to solve the over-smoothing effect of relearning by means of automatic selection of training samples with low classification certainties, which always distribute in or near the edge areas. Finally, EMRF flexibly combines the strengths of relearning and tri-training via the classification certainties calculated by the probabilistic output of the respective classifiers. It should be noted that, in order to achieve an unbiased evaluation, we assessed the classification accuracy of the proposed framework using both edge and non-edge test samples. The experimental results obtained with four multispectral high-resolution images confirm the efficacy of the proposed framework, in terms of both edge and non-edge accuracy.

AB - In recent years, the availability of high-resolution imagery has enabled more detailed observation of the Earth. However, it is imperative to simultaneously achieve accurate interpretation and preserve the spatial details for the classification of such high-resolution data. To this aim, we propose the edge-preservation multi-classifier relearning framework (EMRF). This multi-classifier framework is made up of support vector machine (SVM), random forest (RF), and sparse multinomial logistic regression via variable splitting and augmented Lagrangian (LORSAL) classifiers, considering their complementary characteristics. To better characterize complex scenes of remote sensing images, relearning based on landscape metrics is proposed, which iteratively quantizes both the landscape composition and spatial configuration by the use of the initial classification results. In addition, a novel tri-training strategy is proposed to solve the over-smoothing effect of relearning by means of automatic selection of training samples with low classification certainties, which always distribute in or near the edge areas. Finally, EMRF flexibly combines the strengths of relearning and tri-training via the classification certainties calculated by the probabilistic output of the respective classifiers. It should be noted that, in order to achieve an unbiased evaluation, we assessed the classification accuracy of the proposed framework using both edge and non-edge test samples. The experimental results obtained with four multispectral high-resolution images confirm the efficacy of the proposed framework, in terms of both edge and non-edge accuracy.

KW - ITC-ISI-JOURNAL-ARTICLE

UR - https://ezproxy2.utwente.nl/login?url=https://doi.org/10.1016/j.isprsjprs.2018.02.009

UR - https://ezproxy2.utwente.nl/login?url=https://webapps.itc.utwente.nl/library/2018/isi/yang_edg.pdf

U2 - 10.1016/j.isprsjprs.2018.02.009

DO - 10.1016/j.isprsjprs.2018.02.009

M3 - Article

VL - 138

SP - 57

EP - 73

JO - ISPRS journal of photogrammetry and remote sensing

JF - ISPRS journal of photogrammetry and remote sensing

SN - 0924-2716

ER -