Improved image quilting

  • Authors:
  • Jeremy Long;David Mould

  • Affiliations:
  • University of Saskatchewan;University of Saskatchewan

  • Venue:
  • GI '07 Proceedings of Graphics Interface 2007
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper, we present an improvement to the minimum error boundary cut, a method of shaping texture patches for non-parametric texture synthesis from example algorithms such as Efros and Freeman's Image Quilting [4]. Our method uses an alternate distance metric for Dijkstra's algorithm [3], and as a result we are able to prevent the path from taking short cuts through high cost areas, as can sometimes be seen in traditional image quilting. Furthermore, our method is able to reduce both the maximum error in the resulting texture and the visibility of the remaining defects by spreading them over a longer path. Post-process methods such as pixel re-synthesis [9] can easily be modified and applied to our minimum boundary cut to increase the quality of the results.