An application of statistical methods to determine the appropriate size and location of classification reference area

Kang Ming Lu, Hsien-Te Lin, Chen Yi Sun

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

2 Citations (Scopus)

Abstract

Image classification techniques are usually applied to land-use and land-cover researches. One of the most common means of expressing classification accuracy is the preparation of a classification error matrix. Error matrixes compare the relationship between known reference area data and the corresponding results of a classification. Both the representativeness of reference area data and the classification technique influence the overall accuracy. Representativeness of reference area data means the similar component proportion to study area data. Therefore, the purpose of this paper is to find out statistic methods to assist in determining the most representative reference area. In order to assess different methods to determine reference area we tried to make a simulated urban ground truth data which includes 6 categories: road, building, bare land, vegetation, water, and shadow by supervisors Maximum-likelihood method. We tried to apply the Chi-square goodness of fit test to assess the representativeness of reference areas with different sizes and locations, and calculated the indexes x (Chi-square), P G (the probability of Chi-square cumulative distribution function), and AA (absolute error of classification overall accuracy). The results indicated that the lower x or the higher P G presents higher representative of reference area, and there is a negative functional relationship between x and P G. These results led to the conclusion that the appropriate size and location of the reference area could be determined exactly and efficiently by the index P G.

Original languageEnglish
Title of host publicationAmerican Society for Photogrammetry and Remote Sensing Annual Conference 2009, ASPRS 2009
Pages512-521
Number of pages10
Volume2
Publication statusPublished - 2009
EventAmerican Society for Photogrammetry and Remote Sensing Annual Conference 2009, ASPRS 2009 - Baltimore, MD, United States
Duration: 2009 Mar 92009 Mar 13

Other

OtherAmerican Society for Photogrammetry and Remote Sensing Annual Conference 2009, ASPRS 2009
CountryUnited States
CityBaltimore, MD
Period09-03-0909-03-13

Fingerprint

Statistical methods
matrix
Image classification
Supervisory personnel
image classification
Land use
Maximum likelihood
Distribution functions
land cover
method
Statistics
road
land use
vegetation
Water
water
index

All Science Journal Classification (ASJC) codes

  • Information Systems
  • Computers in Earth Sciences

Cite this

Lu, K. M., Lin, H-T., & Sun, C. Y. (2009). An application of statistical methods to determine the appropriate size and location of classification reference area. In American Society for Photogrammetry and Remote Sensing Annual Conference 2009, ASPRS 2009 (Vol. 2, pp. 512-521)
Lu, Kang Ming ; Lin, Hsien-Te ; Sun, Chen Yi. / An application of statistical methods to determine the appropriate size and location of classification reference area. American Society for Photogrammetry and Remote Sensing Annual Conference 2009, ASPRS 2009. Vol. 2 2009. pp. 512-521
@inproceedings{5d8886779e6c420d9b3fd95783787224,
title = "An application of statistical methods to determine the appropriate size and location of classification reference area",
abstract = "Image classification techniques are usually applied to land-use and land-cover researches. One of the most common means of expressing classification accuracy is the preparation of a classification error matrix. Error matrixes compare the relationship between known reference area data and the corresponding results of a classification. Both the representativeness of reference area data and the classification technique influence the overall accuracy. Representativeness of reference area data means the similar component proportion to study area data. Therefore, the purpose of this paper is to find out statistic methods to assist in determining the most representative reference area. In order to assess different methods to determine reference area we tried to make a simulated urban ground truth data which includes 6 categories: road, building, bare land, vegetation, water, and shadow by supervisors Maximum-likelihood method. We tried to apply the Chi-square goodness of fit test to assess the representativeness of reference areas with different sizes and locations, and calculated the indexes x (Chi-square), P G (the probability of Chi-square cumulative distribution function), and AA (absolute error of classification overall accuracy). The results indicated that the lower x or the higher P G presents higher representative of reference area, and there is a negative functional relationship between x and P G. These results led to the conclusion that the appropriate size and location of the reference area could be determined exactly and efficiently by the index P G.",
author = "Lu, {Kang Ming} and Hsien-Te Lin and Sun, {Chen Yi}",
year = "2009",
language = "English",
isbn = "9781615673223",
volume = "2",
pages = "512--521",
booktitle = "American Society for Photogrammetry and Remote Sensing Annual Conference 2009, ASPRS 2009",

}

Lu, KM, Lin, H-T & Sun, CY 2009, An application of statistical methods to determine the appropriate size and location of classification reference area. in American Society for Photogrammetry and Remote Sensing Annual Conference 2009, ASPRS 2009. vol. 2, pp. 512-521, American Society for Photogrammetry and Remote Sensing Annual Conference 2009, ASPRS 2009, Baltimore, MD, United States, 09-03-09.

An application of statistical methods to determine the appropriate size and location of classification reference area. / Lu, Kang Ming; Lin, Hsien-Te; Sun, Chen Yi.

American Society for Photogrammetry and Remote Sensing Annual Conference 2009, ASPRS 2009. Vol. 2 2009. p. 512-521.

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

TY - GEN

T1 - An application of statistical methods to determine the appropriate size and location of classification reference area

AU - Lu, Kang Ming

AU - Lin, Hsien-Te

AU - Sun, Chen Yi

PY - 2009

Y1 - 2009

N2 - Image classification techniques are usually applied to land-use and land-cover researches. One of the most common means of expressing classification accuracy is the preparation of a classification error matrix. Error matrixes compare the relationship between known reference area data and the corresponding results of a classification. Both the representativeness of reference area data and the classification technique influence the overall accuracy. Representativeness of reference area data means the similar component proportion to study area data. Therefore, the purpose of this paper is to find out statistic methods to assist in determining the most representative reference area. In order to assess different methods to determine reference area we tried to make a simulated urban ground truth data which includes 6 categories: road, building, bare land, vegetation, water, and shadow by supervisors Maximum-likelihood method. We tried to apply the Chi-square goodness of fit test to assess the representativeness of reference areas with different sizes and locations, and calculated the indexes x (Chi-square), P G (the probability of Chi-square cumulative distribution function), and AA (absolute error of classification overall accuracy). The results indicated that the lower x or the higher P G presents higher representative of reference area, and there is a negative functional relationship between x and P G. These results led to the conclusion that the appropriate size and location of the reference area could be determined exactly and efficiently by the index P G.

AB - Image classification techniques are usually applied to land-use and land-cover researches. One of the most common means of expressing classification accuracy is the preparation of a classification error matrix. Error matrixes compare the relationship between known reference area data and the corresponding results of a classification. Both the representativeness of reference area data and the classification technique influence the overall accuracy. Representativeness of reference area data means the similar component proportion to study area data. Therefore, the purpose of this paper is to find out statistic methods to assist in determining the most representative reference area. In order to assess different methods to determine reference area we tried to make a simulated urban ground truth data which includes 6 categories: road, building, bare land, vegetation, water, and shadow by supervisors Maximum-likelihood method. We tried to apply the Chi-square goodness of fit test to assess the representativeness of reference areas with different sizes and locations, and calculated the indexes x (Chi-square), P G (the probability of Chi-square cumulative distribution function), and AA (absolute error of classification overall accuracy). The results indicated that the lower x or the higher P G presents higher representative of reference area, and there is a negative functional relationship between x and P G. These results led to the conclusion that the appropriate size and location of the reference area could be determined exactly and efficiently by the index P G.

UR - http://www.scopus.com/inward/record.url?scp=84868523527&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84868523527&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9781615673223

VL - 2

SP - 512

EP - 521

BT - American Society for Photogrammetry and Remote Sensing Annual Conference 2009, ASPRS 2009

ER -

Lu KM, Lin H-T, Sun CY. An application of statistical methods to determine the appropriate size and location of classification reference area. In American Society for Photogrammetry and Remote Sensing Annual Conference 2009, ASPRS 2009. Vol. 2. 2009. p. 512-521