Image Sequence Analysis via Partial Differential Equations

  • Authors:

  • Pierre Kornprobst;Rachid Deriche;Gilles Aubert

  • Affiliations:

  • INRIA, 2004 route des Lucioles, BP 93, 06902 Sophia-Antipolis Cedex, France. Pierre.Kornprobst@sophia.inria.fr;INRIA, 2004 route des Lucioles, BP 93, 06902 Sophia-Antipolis Cedex, France. Rachid.Deriche@sophia.inria.fr;Laboratoire J.A Dieudonne, UMR no 6621 du CNRS, 06108 Nice-Cedex 2, France. gaubert@math.unice.fr

  • Venue:
  • Journal of Mathematical Imaging and Vision
  • Year:
  • 1999

Quantified Score

Hi-index 0.02

Visualization

Abstract

This article deals with the problem of restoring andmotion segmenting noisy image sequences with a static background.Usually, motion segmentation and image restoration are consideredseparately in image sequence restoration. Moreover, motionsegmentation is often noise sensitive. In this article, the motionsegmentation and the image restoration parts are performed in acoupled way, allowing the motion segmentation part to positivelyinfluence the restoration part and vice-versa. This is the key of ourapproach that allows to deal simultaneously with the problem ofrestoration and motion segmentation. To this end, we propose atheoretically justified optimization problem that permits to takeinto account both requirements. The model is theoreticallyjustified. Existence and unicity are proved in the space of boundedvariations. A suitable numerical scheme based on half quadraticminimization is then proposed and its convergence and stabilitydemonstrated. Experimental results obtained on noisy synthetic dataand real images will illustrate the capabilities of this original andpromising approach.