A distance metric for multidimensional histograms.

M. Werman, S. Peleg, A. Rosenfeld

Research output: Contribution to journalArticlepeer-review

123 Scopus citations

Abstract

A metric is defined on the space of multidimensional histograms. Given two multidimensional histograms, each is 'unfolded' and a minimum distance pairing is performed. The sum of the distances in the minimal pairing is used as the 'match distance' between the histograms. This distance is metric, and in the 1-D case is equal to the absolute difference of the two cumulative distribution functions. It facilitates direct computation of the distance between co-occurrence matrices or between point patterns.-after Authors

Original languageAmerican English
Pages (from-to)328-336
Number of pages9
JournalComputer Vision, Graphics, and Image Processing
Volume32
Issue number3
DOIs
StatePublished - 1985

Bibliographical note

Funding Information:
*Permanent Address: Dept. of Computer Science, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel. These authors were supported in Israel by a grant from the Israel Academy of Sciences. +The support of the National Science Foundation under grant DCR-82-18408 is gratefully acknowledged, as is the help of Janet Salzman in preparing this paper. 328

Fingerprint

Dive into the research topics of 'A distance metric for multidimensional histograms.'. Together they form a unique fingerprint.

Cite this