Efficient evaluations of edge connectivity and width uniformity

Research output: Contribution to journalArticlepeer-review

48 Scopus citations

Abstract

We present three simple and efficient methods for the evaluation of edge connectivity and width uniformity in the images processed by certain edge detection and/or edge linking operations. Quantitative measurements of the edge merits are derived first from an analysis of the edge and pixel patterns in the images, and then extended to two more simplified forms. The main purpose of the evaluation methods is to provide an objective assessment of the qualities of the edge images in terms of their easiness to the successive image processing and vision tasks. The evaluation methods are moderate in terms of the accuracy and sensitivity of the measurements as well as the computational complexity of the related algorithms. No prior knowledge about the true edge positions is needed for deriving the measurement values. The evaluation methods are therefore applicable to the real edge images in practical applications, not only to the synthetic testing images.

Original languageEnglish (US)
Pages (from-to)21-34
Number of pages14
JournalImage and Vision Computing
Volume14
Issue number1
DOIs
StatePublished - Feb 1996

Keywords

  • Edge connectivity
  • Edge quality
  • Width uniformity

ASJC Scopus subject areas

  • Signal Processing
  • Computer Vision and Pattern Recognition

Fingerprint Dive into the research topics of 'Efficient evaluations of edge connectivity and width uniformity'. Together they form a unique fingerprint.

Cite this