Accuracy and inaccuracy assessments in land-cover classification

Ryuei Nishii, Shojiro Tanaka

Research output: Contribution to journalArticle

41 Citations (Scopus)

Abstract

Several measures assessing accuracy of land-cover classification are available, e.g., overall and class-averaged accuracies. Also kappa statistic is widely used for this purpose. In this article, we will discuss properties of these criteria and point out that the kappa statistic has an unfavorable feature. We will propose an alternative coefficient based on Kullback-Leibler information. A test statistic for significance difference between coefficients is also established. Further, the Bayes risk, which takes types of misclassifications into account, is discussed. Our assessment measures are examined through actual error matrices by Ma and Redmond [3].

Original languageEnglish
Pages (from-to)491-498
Number of pages8
JournalIEEE Transactions on Geoscience and Remote Sensing
Volume37
Issue number1 II
DOIs
Publication statusPublished - Dec 1 1999

Fingerprint

land cover
Statistics
matrix
statistics
test

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering
  • Earth and Planetary Sciences(all)

Cite this

Accuracy and inaccuracy assessments in land-cover classification. / Nishii, Ryuei; Tanaka, Shojiro.

In: IEEE Transactions on Geoscience and Remote Sensing, Vol. 37, No. 1 II, 01.12.1999, p. 491-498.

Research output: Contribution to journalArticle

Nishii, Ryuei ; Tanaka, Shojiro. / Accuracy and inaccuracy assessments in land-cover classification. In: IEEE Transactions on Geoscience and Remote Sensing. 1999 ; Vol. 37, No. 1 II. pp. 491-498.
@article{872bec5a8ce84914b1c3011b34ff4966,
title = "Accuracy and inaccuracy assessments in land-cover classification",
abstract = "Several measures assessing accuracy of land-cover classification are available, e.g., overall and class-averaged accuracies. Also kappa statistic is widely used for this purpose. In this article, we will discuss properties of these criteria and point out that the kappa statistic has an unfavorable feature. We will propose an alternative coefficient based on Kullback-Leibler information. A test statistic for significance difference between coefficients is also established. Further, the Bayes risk, which takes types of misclassifications into account, is discussed. Our assessment measures are examined through actual error matrices by Ma and Redmond [3].",
author = "Ryuei Nishii and Shojiro Tanaka",
year = "1999",
month = "12",
day = "1",
doi = "10.1109/36.739098",
language = "English",
volume = "37",
pages = "491--498",
journal = "IEEE Transactions on Geoscience and Remote Sensing",
issn = "0196-2892",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "1 II",

}

TY - JOUR

T1 - Accuracy and inaccuracy assessments in land-cover classification

AU - Nishii, Ryuei

AU - Tanaka, Shojiro

PY - 1999/12/1

Y1 - 1999/12/1

N2 - Several measures assessing accuracy of land-cover classification are available, e.g., overall and class-averaged accuracies. Also kappa statistic is widely used for this purpose. In this article, we will discuss properties of these criteria and point out that the kappa statistic has an unfavorable feature. We will propose an alternative coefficient based on Kullback-Leibler information. A test statistic for significance difference between coefficients is also established. Further, the Bayes risk, which takes types of misclassifications into account, is discussed. Our assessment measures are examined through actual error matrices by Ma and Redmond [3].

AB - Several measures assessing accuracy of land-cover classification are available, e.g., overall and class-averaged accuracies. Also kappa statistic is widely used for this purpose. In this article, we will discuss properties of these criteria and point out that the kappa statistic has an unfavorable feature. We will propose an alternative coefficient based on Kullback-Leibler information. A test statistic for significance difference between coefficients is also established. Further, the Bayes risk, which takes types of misclassifications into account, is discussed. Our assessment measures are examined through actual error matrices by Ma and Redmond [3].

UR - http://www.scopus.com/inward/record.url?scp=0032736309&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0032736309&partnerID=8YFLogxK

U2 - 10.1109/36.739098

DO - 10.1109/36.739098

M3 - Article

VL - 37

SP - 491

EP - 498

JO - IEEE Transactions on Geoscience and Remote Sensing

JF - IEEE Transactions on Geoscience and Remote Sensing

SN - 0196-2892

IS - 1 II

ER -