A self stabilizing robust region finder applied to color and optical flow pictures

M. Ben-Ezra*, M. Werman, Y. Bar-Yam

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

This paper presents a new robust yet very simple algorithm to find a homogeneous region in an image. The algorithm is based on iteratively growing and shrinking a region until a stable state of the algorithm is reached and good fit is found. We have shown that the algorithm works very well in color and optical flow pictures, and overcomes some of the drawbacks of classic seed growing algorithms.

Original languageAmerican English
Pages (from-to)427-433
Number of pages7
JournalImage and Vision Computing
Volume19
Issue number7
DOIs
StatePublished - 1 May 2001

Keywords

  • Region growing
  • Segmentation

Fingerprint

Dive into the research topics of 'A self stabilizing robust region finder applied to color and optical flow pictures'. Together they form a unique fingerprint.

Cite this