Reconstruction of a color image from nonuniformly distributed sparse and noisy data

Dimitris Metaxas, Evangelos Milios

Research output: Contribution to journalArticlepeer-review

7 Scopus citations


This paper presents a physically based approach to the problem of reconstructing a color image from sparse, noisy, and nonuniformly distributed color samples, where spatial sampling does not follow any specific geometry. We reconstruct each of the R, G, and B components of the color image as a thin plate with depth constraints the R, G, and B color samples, respectively. To achieve sharper colors the algorithm uses color discontinuities. We distinguish two cases: First, when color discontinuities are available prior to reconstruction; and second, when they are not a discontinuity detection algorithm is applied as part of the reconstruction process. We present results over a range of sampling densities, including the reconstruction of a color image from 6.25% and 12.5% of the pixels in the first and second case, respectively. We also present results of reconstructing corrupted versions of the original image with zero-mean Gaussian noise. The applicability of the method is independent of the choice of color space.

Original languageEnglish (US)
Pages (from-to)103-111
Number of pages9
JournalCVGIP: Graphical Models and Image Processing
Issue number2
StatePublished - Mar 1992
Externally publishedYes

All Science Journal Classification (ASJC) codes

  • Environmental Science(all)
  • Engineering(all)
  • Earth and Planetary Sciences(all)

Fingerprint Dive into the research topics of 'Reconstruction of a color image from nonuniformly distributed sparse and noisy data'. Together they form a unique fingerprint.

Cite this