An Approximate Distribution for the Normalized Cut

  • Authors:
  • Saralees Nadarajah

  • Affiliations:
  • University of Manchester, Manchester, UK M13 9PL

  • Venue:
  • Journal of Mathematical Imaging and Vision
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

The normalized cut is a popular graph partitioning measure for perceptual organization. Here, some approximate but explicit expressions are derived for the probability density function, cumulative distribution function and the moments of the normalized cut. A simple procedure is provided for computing the associated percentile points and hence the associated confidence intervals. Finally, an application is illustrated.