Variational restoration of nonflat image features: Models and algorithms

Tony Chan*, Jianhong Shen

*Corresponding author for this work

Research output: Contribution to journalReview articlepeer-review

105 Scopus citations

Abstract

We develop both mathematical models and computational algorithms for variational denoising and restoration of nonflat image features. Nonflat image features are those that live on Riemannian manifolds, instead of on the Euclidean spaces. Familiar examples include the orientation feature (from optical flows or gradient flows) that lives on the unit circle S1, the alignment feature (from fingerprint waves or certain texture images) that lives on the real projective line ℝℙ1, and the chromaticity feature (from color images) that lives on the unit sphere S2. In this paper, we apply the variational method to denoise and restore general nonflat image features. Mathematical models for both continuous image domains and discrete domains (or graphs) are constructed. Riemannian objects such as metric, distance and Levi-Civita connection play important roles in the models. Computational algorithms are also developed for the resulting nonlinear equations. The mathematical framework can be applied to restoring general nonflat data outside the scope of image processing and computer vision.

Original languageEnglish (US)
Pages (from-to)1338-1361
Number of pages24
JournalSIAM Journal on Applied Mathematics
Volume61
Issue number4
StatePublished - Dec 1 2000

Keywords

  • Chromaticity
  • Denoising and restoration
  • Metric and distance
  • Nonflat features
  • Orientation, alignment
  • Riemannian manifold
  • Total variation
  • Variational model

ASJC Scopus subject areas

  • Applied Mathematics

Fingerprint Dive into the research topics of 'Variational restoration of nonflat image features: Models and algorithms'. Together they form a unique fingerprint.

Cite this