Spatio-temporal contextual image classification based on spatial AdaBoost

Ryuei Nishii, Shinto Eguchi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

Spatial AdaBoost proposed by Nishii and Eguchi (TGRS 2005) is a contextual supervised classifier of land-cover categories of geostatistical data. It shows an excellent performance similar to that of the MRF-based classifier with much less computational cost. In this paper, we extend the method to the setup with multi spatio-temporal images. We take classification functions by the averages of log posterior probabilities derived by respective training data sets. The functions are sequentially combined by minimizing the empirical exponential risk calculated over samples in all the training data sets. Thus, we obtain a classifier based on a convex combination of the functions. The proposed method is applied to artificial data, and it shows performance similar to that of Spatial AdaBoost based on much larger training data.

Original languageEnglish
Title of host publication25th Anniversary IGARSS 2005: IEEE International Geoscience and Remote Sensing Symposium
Pages172-175
Number of pages4
Volume1
DOIs
Publication statusPublished - 2005
Event2005 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2005 - Seoul, Korea, Republic of
Duration: Jul 25 2005Jul 29 2005

Other

Other2005 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2005
CountryKorea, Republic of
CitySeoul
Period7/25/057/29/05

Fingerprint

Adaptive boosting
Image classification
image classification
Classifiers
land cover
cost
Costs
method

All Science Journal Classification (ASJC) codes

  • Geology
  • Software

Cite this

Nishii, R., & Eguchi, S. (2005). Spatio-temporal contextual image classification based on spatial AdaBoost. In 25th Anniversary IGARSS 2005: IEEE International Geoscience and Remote Sensing Symposium (Vol. 1, pp. 172-175). [1526132] https://doi.org/10.1109/IGARSS.2005.1526132

Spatio-temporal contextual image classification based on spatial AdaBoost. / Nishii, Ryuei; Eguchi, Shinto.

25th Anniversary IGARSS 2005: IEEE International Geoscience and Remote Sensing Symposium. Vol. 1 2005. p. 172-175 1526132.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Nishii, R & Eguchi, S 2005, Spatio-temporal contextual image classification based on spatial AdaBoost. in 25th Anniversary IGARSS 2005: IEEE International Geoscience and Remote Sensing Symposium. vol. 1, 1526132, pp. 172-175, 2005 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2005, Seoul, Korea, Republic of, 7/25/05. https://doi.org/10.1109/IGARSS.2005.1526132
Nishii R, Eguchi S. Spatio-temporal contextual image classification based on spatial AdaBoost. In 25th Anniversary IGARSS 2005: IEEE International Geoscience and Remote Sensing Symposium. Vol. 1. 2005. p. 172-175. 1526132 https://doi.org/10.1109/IGARSS.2005.1526132
Nishii, Ryuei ; Eguchi, Shinto. / Spatio-temporal contextual image classification based on spatial AdaBoost. 25th Anniversary IGARSS 2005: IEEE International Geoscience and Remote Sensing Symposium. Vol. 1 2005. pp. 172-175
@inproceedings{4a6bd9c981b048c28b98beebeb198529,
title = "Spatio-temporal contextual image classification based on spatial AdaBoost",
abstract = "Spatial AdaBoost proposed by Nishii and Eguchi (TGRS 2005) is a contextual supervised classifier of land-cover categories of geostatistical data. It shows an excellent performance similar to that of the MRF-based classifier with much less computational cost. In this paper, we extend the method to the setup with multi spatio-temporal images. We take classification functions by the averages of log posterior probabilities derived by respective training data sets. The functions are sequentially combined by minimizing the empirical exponential risk calculated over samples in all the training data sets. Thus, we obtain a classifier based on a convex combination of the functions. The proposed method is applied to artificial data, and it shows performance similar to that of Spatial AdaBoost based on much larger training data.",
author = "Ryuei Nishii and Shinto Eguchi",
year = "2005",
doi = "10.1109/IGARSS.2005.1526132",
language = "English",
isbn = "0780390504",
volume = "1",
pages = "172--175",
booktitle = "25th Anniversary IGARSS 2005: IEEE International Geoscience and Remote Sensing Symposium",

}

TY - GEN

T1 - Spatio-temporal contextual image classification based on spatial AdaBoost

AU - Nishii, Ryuei

AU - Eguchi, Shinto

PY - 2005

Y1 - 2005

N2 - Spatial AdaBoost proposed by Nishii and Eguchi (TGRS 2005) is a contextual supervised classifier of land-cover categories of geostatistical data. It shows an excellent performance similar to that of the MRF-based classifier with much less computational cost. In this paper, we extend the method to the setup with multi spatio-temporal images. We take classification functions by the averages of log posterior probabilities derived by respective training data sets. The functions are sequentially combined by minimizing the empirical exponential risk calculated over samples in all the training data sets. Thus, we obtain a classifier based on a convex combination of the functions. The proposed method is applied to artificial data, and it shows performance similar to that of Spatial AdaBoost based on much larger training data.

AB - Spatial AdaBoost proposed by Nishii and Eguchi (TGRS 2005) is a contextual supervised classifier of land-cover categories of geostatistical data. It shows an excellent performance similar to that of the MRF-based classifier with much less computational cost. In this paper, we extend the method to the setup with multi spatio-temporal images. We take classification functions by the averages of log posterior probabilities derived by respective training data sets. The functions are sequentially combined by minimizing the empirical exponential risk calculated over samples in all the training data sets. Thus, we obtain a classifier based on a convex combination of the functions. The proposed method is applied to artificial data, and it shows performance similar to that of Spatial AdaBoost based on much larger training data.

UR - http://www.scopus.com/inward/record.url?scp=33745699902&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33745699902&partnerID=8YFLogxK

U2 - 10.1109/IGARSS.2005.1526132

DO - 10.1109/IGARSS.2005.1526132

M3 - Conference contribution

AN - SCOPUS:33745699902

SN - 0780390504

SN - 9780780390508

VL - 1

SP - 172

EP - 175

BT - 25th Anniversary IGARSS 2005: IEEE International Geoscience and Remote Sensing Symposium

ER -