Per-pixel vs. object-based classification of urban land cover extraction using high spatial resolution imagery

Soe Myint, Patricia Gober, Anthony Brazel, Susanne Grossman-Clarke, Qihao Weng

Research output: Contribution to journalArticle

651 Citations (Scopus)

Abstract

In using traditional digital classification algorithms, a researcher typically encounters serious issues in identifying urban land cover classes employing high resolution data. A normal approach is to use spectral information alone and ignore spatial information and a group of pixels that need to be considered together as an object. We used QuickBird image data over a central region in the city of Phoenix, Arizona to examine if an object-based classifier can accurately identify urban classes. To demonstrate if spectral information alone is practical in urban classification, we used spectra of the selected classes from randomly selected points to examine if they can be effectively discriminated. The overall accuracy based on spectral information alone reached only about 63.33%. We employed five different classification procedures with the object-based paradigm that separates spatially and spectrally similar pixels at different scales. The classifiers to assign land covers to segmented objects used in the study include membership functions and the nearest neighbor classifier. The object-based classifier achieved a high overall accuracy (90.40%), whereas the most commonly used decision rule, namely maximum likelihood classifier, produced a lower overall accuracy (67.60%). This study demonstrates that the object-based classifier is a significantly better approach than the classical per-pixel classifiers. Further, this study reviews application of different parameters for segmentation and classification, combined use of composite and original bands, selection of different scale levels, and choice of classifiers. Strengths and weaknesses of the object-based prototype are presented and we provide suggestions to avoid or minimize uncertainties and limitations associated with the approach.

Original languageEnglish (US)
Pages (from-to)1145-1161
Number of pages17
JournalRemote Sensing of Environment
Volume115
Issue number5
DOIs
StatePublished - May 15 2011

Fingerprint

land cover
pixel
spatial resolution
Classifiers
imagery
Pixels
application parameters
QuickBird
prototypes
segmentation
uncertainty
researchers
Membership functions
Maximum likelihood
Composite materials

Keywords

  • High resolution
  • Membership function
  • Nearest neighbor
  • Object-based classifier
  • Urban

ASJC Scopus subject areas

  • Computers in Earth Sciences
  • Soil Science
  • Geology

Cite this

Per-pixel vs. object-based classification of urban land cover extraction using high spatial resolution imagery. / Myint, Soe; Gober, Patricia; Brazel, Anthony; Grossman-Clarke, Susanne; Weng, Qihao.

In: Remote Sensing of Environment, Vol. 115, No. 5, 15.05.2011, p. 1145-1161.

Research output: Contribution to journalArticle

Myint, Soe ; Gober, Patricia ; Brazel, Anthony ; Grossman-Clarke, Susanne ; Weng, Qihao. / Per-pixel vs. object-based classification of urban land cover extraction using high spatial resolution imagery. In: Remote Sensing of Environment. 2011 ; Vol. 115, No. 5. pp. 1145-1161.
@article{a54e64a6b55543fc842576f7ff1f3418,
title = "Per-pixel vs. object-based classification of urban land cover extraction using high spatial resolution imagery",
abstract = "In using traditional digital classification algorithms, a researcher typically encounters serious issues in identifying urban land cover classes employing high resolution data. A normal approach is to use spectral information alone and ignore spatial information and a group of pixels that need to be considered together as an object. We used QuickBird image data over a central region in the city of Phoenix, Arizona to examine if an object-based classifier can accurately identify urban classes. To demonstrate if spectral information alone is practical in urban classification, we used spectra of the selected classes from randomly selected points to examine if they can be effectively discriminated. The overall accuracy based on spectral information alone reached only about 63.33{\%}. We employed five different classification procedures with the object-based paradigm that separates spatially and spectrally similar pixels at different scales. The classifiers to assign land covers to segmented objects used in the study include membership functions and the nearest neighbor classifier. The object-based classifier achieved a high overall accuracy (90.40{\%}), whereas the most commonly used decision rule, namely maximum likelihood classifier, produced a lower overall accuracy (67.60{\%}). This study demonstrates that the object-based classifier is a significantly better approach than the classical per-pixel classifiers. Further, this study reviews application of different parameters for segmentation and classification, combined use of composite and original bands, selection of different scale levels, and choice of classifiers. Strengths and weaknesses of the object-based prototype are presented and we provide suggestions to avoid or minimize uncertainties and limitations associated with the approach.",
keywords = "High resolution, Membership function, Nearest neighbor, Object-based classifier, Urban",
author = "Soe Myint and Patricia Gober and Anthony Brazel and Susanne Grossman-Clarke and Qihao Weng",
year = "2011",
month = "5",
day = "15",
doi = "10.1016/j.rse.2010.12.017",
language = "English (US)",
volume = "115",
pages = "1145--1161",
journal = "Remote Sensing of Environment",
issn = "0034-4257",
publisher = "Elsevier Inc.",
number = "5",

}

TY - JOUR

T1 - Per-pixel vs. object-based classification of urban land cover extraction using high spatial resolution imagery

AU - Myint, Soe

AU - Gober, Patricia

AU - Brazel, Anthony

AU - Grossman-Clarke, Susanne

AU - Weng, Qihao

PY - 2011/5/15

Y1 - 2011/5/15

N2 - In using traditional digital classification algorithms, a researcher typically encounters serious issues in identifying urban land cover classes employing high resolution data. A normal approach is to use spectral information alone and ignore spatial information and a group of pixels that need to be considered together as an object. We used QuickBird image data over a central region in the city of Phoenix, Arizona to examine if an object-based classifier can accurately identify urban classes. To demonstrate if spectral information alone is practical in urban classification, we used spectra of the selected classes from randomly selected points to examine if they can be effectively discriminated. The overall accuracy based on spectral information alone reached only about 63.33%. We employed five different classification procedures with the object-based paradigm that separates spatially and spectrally similar pixels at different scales. The classifiers to assign land covers to segmented objects used in the study include membership functions and the nearest neighbor classifier. The object-based classifier achieved a high overall accuracy (90.40%), whereas the most commonly used decision rule, namely maximum likelihood classifier, produced a lower overall accuracy (67.60%). This study demonstrates that the object-based classifier is a significantly better approach than the classical per-pixel classifiers. Further, this study reviews application of different parameters for segmentation and classification, combined use of composite and original bands, selection of different scale levels, and choice of classifiers. Strengths and weaknesses of the object-based prototype are presented and we provide suggestions to avoid or minimize uncertainties and limitations associated with the approach.

AB - In using traditional digital classification algorithms, a researcher typically encounters serious issues in identifying urban land cover classes employing high resolution data. A normal approach is to use spectral information alone and ignore spatial information and a group of pixels that need to be considered together as an object. We used QuickBird image data over a central region in the city of Phoenix, Arizona to examine if an object-based classifier can accurately identify urban classes. To demonstrate if spectral information alone is practical in urban classification, we used spectra of the selected classes from randomly selected points to examine if they can be effectively discriminated. The overall accuracy based on spectral information alone reached only about 63.33%. We employed five different classification procedures with the object-based paradigm that separates spatially and spectrally similar pixels at different scales. The classifiers to assign land covers to segmented objects used in the study include membership functions and the nearest neighbor classifier. The object-based classifier achieved a high overall accuracy (90.40%), whereas the most commonly used decision rule, namely maximum likelihood classifier, produced a lower overall accuracy (67.60%). This study demonstrates that the object-based classifier is a significantly better approach than the classical per-pixel classifiers. Further, this study reviews application of different parameters for segmentation and classification, combined use of composite and original bands, selection of different scale levels, and choice of classifiers. Strengths and weaknesses of the object-based prototype are presented and we provide suggestions to avoid or minimize uncertainties and limitations associated with the approach.

KW - High resolution

KW - Membership function

KW - Nearest neighbor

KW - Object-based classifier

KW - Urban

UR - http://www.scopus.com/inward/record.url?scp=79952070569&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79952070569&partnerID=8YFLogxK

U2 - 10.1016/j.rse.2010.12.017

DO - 10.1016/j.rse.2010.12.017

M3 - Article

AN - SCOPUS:79952070569

VL - 115

SP - 1145

EP - 1161

JO - Remote Sensing of Environment

JF - Remote Sensing of Environment

SN - 0034-4257

IS - 5

ER -