Dense Non-Rigid Motion Estimation in Sequencesof Medical Images Using Differential Constraints

  • Authors:
  • Serge Benayoun;Nicholas Ayache

  • Affiliations:
  • INRIA, B.P. 93, 06902 Sophia-Antipolis Cedex, France;INRIA, B.P. 93, 06902 Sophia-Antipolis Cedex, France

  • Venue:
  • International Journal of Computer Vision
  • Year:
  • 1998

Quantified Score

Hi-index 0.00

Visualization

Abstract

We describe a new method for computing the displacement vector field in time sequences of2D or 3D images (4D data). The method is energy-minimizing on the space ofcorrespondence functions; the energy is split into twoterms, with one term matchingdifferential singularities in the images, and the other constraining the regularity of thefield. In order to reduce the computational timeof the motion estimation, we use an adaptive image mesh, the resolution of which depends onthe value of the gradient intensity. We solve numerically theminimization problem with thefinite element method which gives a continuous approximation of the solution. We presentexperimental results on synthetic data and on medical images and we show how to use these results for analyzing cardiac deformations.