A similarity metric for edge images

M. S. Prieto, Alastair Robert Allen

Research output: Contribution to journalArticle

75 Citations (Scopus)

Abstract

The performance of several discrepancy measures for the comparison of edge images is analyzed and a novel similarity metric aimed at overcoming their problems is proposed. The algorithm finds an optimal matching of the pixels between the images and estimates the error produced by this matching. The resulting Pixel Correspondence Metric (PCM) can take into account edge strength as well as the displacement of edge pixel positions in the estimation of similarity. A series of experimental tests shows the new metric to be a robust and effective tool in the comparison of edge images when a small localization error of the detected edges is allowed.

Original languageEnglish
Pages (from-to)1265-1273
Number of pages8
JournalIEEE Transactions on Pattern Analysis and Machine Intelligence
Volume25
DOIs
Publication statusPublished - 2003

Keywords

  • edge detection
  • image similarity
  • performance evaluation
  • pixel correspondence metric
  • weighted matching in bipartite graphs
  • matching problem
  • assignment problem
  • PERFORMANCE EVALUATION
  • ALGORITHMS

Cite this

A similarity metric for edge images. / Prieto, M. S.; Allen, Alastair Robert.

In: IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 25, 2003, p. 1265-1273.

Research output: Contribution to journalArticle

@article{e3f082b229d747b58441a96d76c8c77b,
title = "A similarity metric for edge images",
abstract = "The performance of several discrepancy measures for the comparison of edge images is analyzed and a novel similarity metric aimed at overcoming their problems is proposed. The algorithm finds an optimal matching of the pixels between the images and estimates the error produced by this matching. The resulting Pixel Correspondence Metric (PCM) can take into account edge strength as well as the displacement of edge pixel positions in the estimation of similarity. A series of experimental tests shows the new metric to be a robust and effective tool in the comparison of edge images when a small localization error of the detected edges is allowed.",
keywords = "edge detection, image similarity, performance evaluation, pixel correspondence metric, weighted matching in bipartite graphs, matching problem, assignment problem, PERFORMANCE EVALUATION, ALGORITHMS",
author = "Prieto, {M. S.} and Allen, {Alastair Robert}",
year = "2003",
doi = "10.1109/TPAMI.2003.1233900",
language = "English",
volume = "25",
pages = "1265--1273",
journal = "IEEE Transactions on Pattern Analysis and Machine Intelligence",
issn = "0162-8828",
publisher = "IEEE Computer Society",

}

TY - JOUR

T1 - A similarity metric for edge images

AU - Prieto, M. S.

AU - Allen, Alastair Robert

PY - 2003

Y1 - 2003

N2 - The performance of several discrepancy measures for the comparison of edge images is analyzed and a novel similarity metric aimed at overcoming their problems is proposed. The algorithm finds an optimal matching of the pixels between the images and estimates the error produced by this matching. The resulting Pixel Correspondence Metric (PCM) can take into account edge strength as well as the displacement of edge pixel positions in the estimation of similarity. A series of experimental tests shows the new metric to be a robust and effective tool in the comparison of edge images when a small localization error of the detected edges is allowed.

AB - The performance of several discrepancy measures for the comparison of edge images is analyzed and a novel similarity metric aimed at overcoming their problems is proposed. The algorithm finds an optimal matching of the pixels between the images and estimates the error produced by this matching. The resulting Pixel Correspondence Metric (PCM) can take into account edge strength as well as the displacement of edge pixel positions in the estimation of similarity. A series of experimental tests shows the new metric to be a robust and effective tool in the comparison of edge images when a small localization error of the detected edges is allowed.

KW - edge detection

KW - image similarity

KW - performance evaluation

KW - pixel correspondence metric

KW - weighted matching in bipartite graphs

KW - matching problem

KW - assignment problem

KW - PERFORMANCE EVALUATION

KW - ALGORITHMS

U2 - 10.1109/TPAMI.2003.1233900

DO - 10.1109/TPAMI.2003.1233900

M3 - Article

VL - 25

SP - 1265

EP - 1273

JO - IEEE Transactions on Pattern Analysis and Machine Intelligence

JF - IEEE Transactions on Pattern Analysis and Machine Intelligence

SN - 0162-8828

ER -