Details
| Original language | English |
|---|---|
| Pages (from-to) | 117-122 |
| Number of pages | 6 |
| Journal | ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
| Volume | 1 |
| Publication status | Published - 17 Jul 2012 |
| Event | 22nd Congress of the International Society for Photogrammetry and Remote Sensing: Imaging a Sustainable Future, ISPRS 2012 - Melbourne, Australia Duration: 25 Aug 2012 → 1 Sept 2012 |
Abstract
In this paper, we present a method for automatic refinement of training data. Many classifiers from machine learning used in applications in the remote sensing domain, rely on previously labelled training data. This labelling is often done by human operators and is bound to time constraints. Hence, selection of training data must be kept practical which implies a certain inaccuracy. This results in erroneously tagged regions enclosed within competing classes. For that purpose, we propose a method that removes outliers from training data by using an iterative training-classification scheme. Outliers are detected by their newly determined class membership as well as through analysis of uncertainty of classified samples. The sample selection method which incorporates quality of neighbouring samples is presented and compared to alternative strategies. Additionally, iterative approaches tend to propagate errors which might lead to degenerating classes. Therefore, a robust stopping criterion based on training data characteristics is described. Our experiments using a support vector machine (SVM) show, that outliers are reliably removed, allowing a more convenient sample selection. The classification result for unknown scenes of the accordant validation set improves from 70.36% to 79.12% on average. Additionally, the average complexity of the SVM model is decreased by 82.75% resulting in similar reduction of processing time.
Keywords
- Classification, Imagery, Land Cover, Learning, Satellite, Training
ASJC Scopus subject areas
- Earth and Planetary Sciences(all)
- Earth and Planetary Sciences (miscellaneous)
- Environmental Science(all)
- Environmental Science (miscellaneous)
- Physics and Astronomy(all)
- Instrumentation
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol. 1, 17.07.2012, p. 117-122.
Research output: Contribution to journal › Conference article › Research › peer review
}
TY - JOUR
T1 - Automatic refinement of training data for classification of satellite imagery
AU - Büschenfeld, Torsten
AU - Ostermann, Jörn
PY - 2012/7/17
Y1 - 2012/7/17
N2 - In this paper, we present a method for automatic refinement of training data. Many classifiers from machine learning used in applications in the remote sensing domain, rely on previously labelled training data. This labelling is often done by human operators and is bound to time constraints. Hence, selection of training data must be kept practical which implies a certain inaccuracy. This results in erroneously tagged regions enclosed within competing classes. For that purpose, we propose a method that removes outliers from training data by using an iterative training-classification scheme. Outliers are detected by their newly determined class membership as well as through analysis of uncertainty of classified samples. The sample selection method which incorporates quality of neighbouring samples is presented and compared to alternative strategies. Additionally, iterative approaches tend to propagate errors which might lead to degenerating classes. Therefore, a robust stopping criterion based on training data characteristics is described. Our experiments using a support vector machine (SVM) show, that outliers are reliably removed, allowing a more convenient sample selection. The classification result for unknown scenes of the accordant validation set improves from 70.36% to 79.12% on average. Additionally, the average complexity of the SVM model is decreased by 82.75% resulting in similar reduction of processing time.
AB - In this paper, we present a method for automatic refinement of training data. Many classifiers from machine learning used in applications in the remote sensing domain, rely on previously labelled training data. This labelling is often done by human operators and is bound to time constraints. Hence, selection of training data must be kept practical which implies a certain inaccuracy. This results in erroneously tagged regions enclosed within competing classes. For that purpose, we propose a method that removes outliers from training data by using an iterative training-classification scheme. Outliers are detected by their newly determined class membership as well as through analysis of uncertainty of classified samples. The sample selection method which incorporates quality of neighbouring samples is presented and compared to alternative strategies. Additionally, iterative approaches tend to propagate errors which might lead to degenerating classes. Therefore, a robust stopping criterion based on training data characteristics is described. Our experiments using a support vector machine (SVM) show, that outliers are reliably removed, allowing a more convenient sample selection. The classification result for unknown scenes of the accordant validation set improves from 70.36% to 79.12% on average. Additionally, the average complexity of the SVM model is decreased by 82.75% resulting in similar reduction of processing time.
KW - Classification
KW - Imagery
KW - Land Cover
KW - Learning
KW - Satellite
KW - Training
UR - http://www.scopus.com/inward/record.url?scp=84962306659&partnerID=8YFLogxK
U2 - 10.5194/isprsannals-I-7-117-2012
DO - 10.5194/isprsannals-I-7-117-2012
M3 - Conference article
AN - SCOPUS:84962306659
VL - 1
SP - 117
EP - 122
JO - ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
JF - ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
SN - 2194-9042
T2 - 22nd Congress of the International Society for Photogrammetry and Remote Sensing: Imaging a Sustainable Future, ISPRS 2012
Y2 - 25 August 2012 through 1 September 2012
ER -