Radiometric normalization of multitemporal optical satellite images using iteratively-reweighted multivariate alteration detection

Bo Yi Lin, Zhi Jia Wang, Chao Hung Lin

研究成果: Paper

摘要

Radiometric normalization is a fundamental process for multitemporal satellite images. The accuracy of relative normalization depends on the quality of selected Pseudo Invariant Features (PIFs). PIFs represent the ground objects whose reflectance are constant during a period of time. In previous study, an algorithm, called Multivariate Alteration Detection (MAD), was applied to statistically select no-changed pixels within bi-temporal satellite images. However, MAD is sensitive to cloud covers and some clouds may be misclassified as PIFs. For this reason, Iteratively Reweighted MAD (IR-MAD) was introduced to establish an increasingly better no-changed background using iterative scheme. Nonetheless, both MAD and IR-MAD only compute the linear combinations for bi-temporal images, and not applicable for multitemporal images with more than two images. In this study, a novel method called Weighted Generalized Canonical Correlation Analysis (WGCCA) is proposed for the selection of high-quality PIFs in multitemporal and multispectral images, which solves coefficients for the correlations of not only multivariable data but also multitemporal data. Specifically, the proposed method integrates the strengths of Generalized Canonical Correlation Analysis (GCCA) and IR-MAD, and PIFs extraction from a sequence of satellite images is performed at the same time, which leads to a consistent feature extraction. Furthermore, when the high-quality PIFs are determined by the proposed method, the digital numbers of PIFs from multitemporal images are transformed into a predefined radiometric reference level. With this approach, the radiometric resolution of multitemporal images can be preserved, and a better radiometric normalization can be obtained. In experiment, SPOT-5 imagery was tested. Compared with Canonical Correlation Analysis (CCA) which is used in MAD, the proposed method can discriminate no-changed pixels from changed more precisely.

原文English
出版狀態Published - 2017 一月 1
事件38th Asian Conference on Remote Sensing - Space Applications: Touching Human Lives, ACRS 2017 - New Delhi, India
持續時間: 2017 十月 232017 十月 27

Other

Other38th Asian Conference on Remote Sensing - Space Applications: Touching Human Lives, ACRS 2017
國家India
城市New Delhi
期間17-10-2317-10-27

指紋

Satellites
Feature extraction
Pixels
Experiments

All Science Journal Classification (ASJC) codes

  • Computer Networks and Communications

引用此文

Lin, B. Y., Wang, Z. J., & Lin, C. H. (2017). Radiometric normalization of multitemporal optical satellite images using iteratively-reweighted multivariate alteration detection. 論文發表於 38th Asian Conference on Remote Sensing - Space Applications: Touching Human Lives, ACRS 2017, New Delhi, India.
Lin, Bo Yi ; Wang, Zhi Jia ; Lin, Chao Hung. / Radiometric normalization of multitemporal optical satellite images using iteratively-reweighted multivariate alteration detection. 論文發表於 38th Asian Conference on Remote Sensing - Space Applications: Touching Human Lives, ACRS 2017, New Delhi, India.
@conference{176bdc69154043cb92f454375440acee,
title = "Radiometric normalization of multitemporal optical satellite images using iteratively-reweighted multivariate alteration detection",
abstract = "Radiometric normalization is a fundamental process for multitemporal satellite images. The accuracy of relative normalization depends on the quality of selected Pseudo Invariant Features (PIFs). PIFs represent the ground objects whose reflectance are constant during a period of time. In previous study, an algorithm, called Multivariate Alteration Detection (MAD), was applied to statistically select no-changed pixels within bi-temporal satellite images. However, MAD is sensitive to cloud covers and some clouds may be misclassified as PIFs. For this reason, Iteratively Reweighted MAD (IR-MAD) was introduced to establish an increasingly better no-changed background using iterative scheme. Nonetheless, both MAD and IR-MAD only compute the linear combinations for bi-temporal images, and not applicable for multitemporal images with more than two images. In this study, a novel method called Weighted Generalized Canonical Correlation Analysis (WGCCA) is proposed for the selection of high-quality PIFs in multitemporal and multispectral images, which solves coefficients for the correlations of not only multivariable data but also multitemporal data. Specifically, the proposed method integrates the strengths of Generalized Canonical Correlation Analysis (GCCA) and IR-MAD, and PIFs extraction from a sequence of satellite images is performed at the same time, which leads to a consistent feature extraction. Furthermore, when the high-quality PIFs are determined by the proposed method, the digital numbers of PIFs from multitemporal images are transformed into a predefined radiometric reference level. With this approach, the radiometric resolution of multitemporal images can be preserved, and a better radiometric normalization can be obtained. In experiment, SPOT-5 imagery was tested. Compared with Canonical Correlation Analysis (CCA) which is used in MAD, the proposed method can discriminate no-changed pixels from changed more precisely.",
author = "Lin, {Bo Yi} and Wang, {Zhi Jia} and Lin, {Chao Hung}",
year = "2017",
month = "1",
day = "1",
language = "English",
note = "38th Asian Conference on Remote Sensing - Space Applications: Touching Human Lives, ACRS 2017 ; Conference date: 23-10-2017 Through 27-10-2017",

}

Lin, BY, Wang, ZJ & Lin, CH 2017, 'Radiometric normalization of multitemporal optical satellite images using iteratively-reweighted multivariate alteration detection', 論文發表於 38th Asian Conference on Remote Sensing - Space Applications: Touching Human Lives, ACRS 2017, New Delhi, India, 17-10-23 - 17-10-27.

Radiometric normalization of multitemporal optical satellite images using iteratively-reweighted multivariate alteration detection. / Lin, Bo Yi; Wang, Zhi Jia; Lin, Chao Hung.

2017. 論文發表於 38th Asian Conference on Remote Sensing - Space Applications: Touching Human Lives, ACRS 2017, New Delhi, India.

研究成果: Paper

TY - CONF

T1 - Radiometric normalization of multitemporal optical satellite images using iteratively-reweighted multivariate alteration detection

AU - Lin, Bo Yi

AU - Wang, Zhi Jia

AU - Lin, Chao Hung

PY - 2017/1/1

Y1 - 2017/1/1

N2 - Radiometric normalization is a fundamental process for multitemporal satellite images. The accuracy of relative normalization depends on the quality of selected Pseudo Invariant Features (PIFs). PIFs represent the ground objects whose reflectance are constant during a period of time. In previous study, an algorithm, called Multivariate Alteration Detection (MAD), was applied to statistically select no-changed pixels within bi-temporal satellite images. However, MAD is sensitive to cloud covers and some clouds may be misclassified as PIFs. For this reason, Iteratively Reweighted MAD (IR-MAD) was introduced to establish an increasingly better no-changed background using iterative scheme. Nonetheless, both MAD and IR-MAD only compute the linear combinations for bi-temporal images, and not applicable for multitemporal images with more than two images. In this study, a novel method called Weighted Generalized Canonical Correlation Analysis (WGCCA) is proposed for the selection of high-quality PIFs in multitemporal and multispectral images, which solves coefficients for the correlations of not only multivariable data but also multitemporal data. Specifically, the proposed method integrates the strengths of Generalized Canonical Correlation Analysis (GCCA) and IR-MAD, and PIFs extraction from a sequence of satellite images is performed at the same time, which leads to a consistent feature extraction. Furthermore, when the high-quality PIFs are determined by the proposed method, the digital numbers of PIFs from multitemporal images are transformed into a predefined radiometric reference level. With this approach, the radiometric resolution of multitemporal images can be preserved, and a better radiometric normalization can be obtained. In experiment, SPOT-5 imagery was tested. Compared with Canonical Correlation Analysis (CCA) which is used in MAD, the proposed method can discriminate no-changed pixels from changed more precisely.

AB - Radiometric normalization is a fundamental process for multitemporal satellite images. The accuracy of relative normalization depends on the quality of selected Pseudo Invariant Features (PIFs). PIFs represent the ground objects whose reflectance are constant during a period of time. In previous study, an algorithm, called Multivariate Alteration Detection (MAD), was applied to statistically select no-changed pixels within bi-temporal satellite images. However, MAD is sensitive to cloud covers and some clouds may be misclassified as PIFs. For this reason, Iteratively Reweighted MAD (IR-MAD) was introduced to establish an increasingly better no-changed background using iterative scheme. Nonetheless, both MAD and IR-MAD only compute the linear combinations for bi-temporal images, and not applicable for multitemporal images with more than two images. In this study, a novel method called Weighted Generalized Canonical Correlation Analysis (WGCCA) is proposed for the selection of high-quality PIFs in multitemporal and multispectral images, which solves coefficients for the correlations of not only multivariable data but also multitemporal data. Specifically, the proposed method integrates the strengths of Generalized Canonical Correlation Analysis (GCCA) and IR-MAD, and PIFs extraction from a sequence of satellite images is performed at the same time, which leads to a consistent feature extraction. Furthermore, when the high-quality PIFs are determined by the proposed method, the digital numbers of PIFs from multitemporal images are transformed into a predefined radiometric reference level. With this approach, the radiometric resolution of multitemporal images can be preserved, and a better radiometric normalization can be obtained. In experiment, SPOT-5 imagery was tested. Compared with Canonical Correlation Analysis (CCA) which is used in MAD, the proposed method can discriminate no-changed pixels from changed more precisely.

UR - http://www.scopus.com/inward/record.url?scp=85047403810&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85047403810&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:85047403810

ER -

Lin BY, Wang ZJ, Lin CH. Radiometric normalization of multitemporal optical satellite images using iteratively-reweighted multivariate alteration detection. 2017. 論文發表於 38th Asian Conference on Remote Sensing - Space Applications: Touching Human Lives, ACRS 2017, New Delhi, India.