Error measures for scene segmentation

William A. Yasnoff, Jack K. Mui, James W. Bacus

Research output: Contribution to journalArticle

Abstract

Scene segmentation is an important problem in pattern recognition. Current subjective methods for evaluation and comparison of scene segmentation techniques are inadequate and objective quantitative measures are desirable. Two error measures, the percentage area misclassified (p) and a new pixel distance error (ε{lunate}) were defined and evaluated in terms of their correlation with human observation for comparison of multiple segmentations of the same scene and multiple scenes segmented by the same technique. The results indicate that both these measures can be helpful in the evaluation and comparison of scene segmentation procedures.

Original languageEnglish (US)
Pages (from-to)217-231
Number of pages15
JournalPattern Recognition
Volume9
Issue number4
DOIs
StatePublished - 1977
Externally publishedYes

Fingerprint

Pattern recognition
Pixels

Keywords

  • Blood cells
  • Confusion matrix
  • Error measures
  • Image processing
  • Pixel distance error
  • Psychopictorics
  • Scene segmentation
  • Type I and II error

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Signal Processing
  • Electrical and Electronic Engineering

Cite this

Yasnoff, W. A., Mui, J. K., & Bacus, J. W. (1977). Error measures for scene segmentation. Pattern Recognition, 9(4), 217-231. https://doi.org/10.1016/0031-3203(77)90006-1

Error measures for scene segmentation. / Yasnoff, William A.; Mui, Jack K.; Bacus, James W.

In: Pattern Recognition, Vol. 9, No. 4, 1977, p. 217-231.

Research output: Contribution to journalArticle

Yasnoff, WA, Mui, JK & Bacus, JW 1977, 'Error measures for scene segmentation', Pattern Recognition, vol. 9, no. 4, pp. 217-231. https://doi.org/10.1016/0031-3203(77)90006-1
Yasnoff, William A. ; Mui, Jack K. ; Bacus, James W. / Error measures for scene segmentation. In: Pattern Recognition. 1977 ; Vol. 9, No. 4. pp. 217-231.
@article{1e4b7f0e7ebc4f508be117c3e71c1576,
title = "Error measures for scene segmentation",
abstract = "Scene segmentation is an important problem in pattern recognition. Current subjective methods for evaluation and comparison of scene segmentation techniques are inadequate and objective quantitative measures are desirable. Two error measures, the percentage area misclassified (p) and a new pixel distance error (ε{lunate}) were defined and evaluated in terms of their correlation with human observation for comparison of multiple segmentations of the same scene and multiple scenes segmented by the same technique. The results indicate that both these measures can be helpful in the evaluation and comparison of scene segmentation procedures.",
keywords = "Blood cells, Confusion matrix, Error measures, Image processing, Pixel distance error, Psychopictorics, Scene segmentation, Type I and II error",
author = "Yasnoff, {William A.} and Mui, {Jack K.} and Bacus, {James W.}",
year = "1977",
doi = "10.1016/0031-3203(77)90006-1",
language = "English (US)",
volume = "9",
pages = "217--231",
journal = "Pattern Recognition",
issn = "0031-3203",
publisher = "Elsevier Limited",
number = "4",

}

TY - JOUR

T1 - Error measures for scene segmentation

AU - Yasnoff, William A.

AU - Mui, Jack K.

AU - Bacus, James W.

PY - 1977

Y1 - 1977

N2 - Scene segmentation is an important problem in pattern recognition. Current subjective methods for evaluation and comparison of scene segmentation techniques are inadequate and objective quantitative measures are desirable. Two error measures, the percentage area misclassified (p) and a new pixel distance error (ε{lunate}) were defined and evaluated in terms of their correlation with human observation for comparison of multiple segmentations of the same scene and multiple scenes segmented by the same technique. The results indicate that both these measures can be helpful in the evaluation and comparison of scene segmentation procedures.

AB - Scene segmentation is an important problem in pattern recognition. Current subjective methods for evaluation and comparison of scene segmentation techniques are inadequate and objective quantitative measures are desirable. Two error measures, the percentage area misclassified (p) and a new pixel distance error (ε{lunate}) were defined and evaluated in terms of their correlation with human observation for comparison of multiple segmentations of the same scene and multiple scenes segmented by the same technique. The results indicate that both these measures can be helpful in the evaluation and comparison of scene segmentation procedures.

KW - Blood cells

KW - Confusion matrix

KW - Error measures

KW - Image processing

KW - Pixel distance error

KW - Psychopictorics

KW - Scene segmentation

KW - Type I and II error

UR - http://www.scopus.com/inward/record.url?scp=0017677110&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0017677110&partnerID=8YFLogxK

U2 - 10.1016/0031-3203(77)90006-1

DO - 10.1016/0031-3203(77)90006-1

M3 - Article

AN - SCOPUS:0017677110

VL - 9

SP - 217

EP - 231

JO - Pattern Recognition

JF - Pattern Recognition

SN - 0031-3203

IS - 4

ER -