Skip to Main content Skip to Navigation
Theses

Recalage de signaux et analyse de variance fonctionnelle par ondelettes. Applications au domaine biomédical

Abstract : This thesis focuses on landmark-based registration of curves and images for the comparison of multiple sets of signals. After a detailed review of some important registration techniques, a nonparametric approach is proposed to estimate the landmarks of a 1D function observed with noise, from the wavelet maxima and the zero-crossings lines of its continuous wavelet transform. A new tool, the structural intensity, is introduced to represent the locations of the typical features of a curve via a probability density function. This method yields to a new technique to automatically put into correspondence the landmarks of two functions. The approach developed for 1D signals is extended on a 2D setting by using wedgelets/platelets decompositions to detect the edges of an image. The usefulness of these methods is documented by various simulations and applications to real examples involving functional analysis of variance.
Complete list of metadata

Cited literature [97 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-00003344
Contributor : Thèses Imag <>
Submitted on : Thursday, January 29, 2004 - 10:12:23 AM
Last modification on : Wednesday, March 10, 2021 - 1:50:03 PM
Long-term archiving on: : Monday, September 20, 2010 - 11:54:48 AM

Identifiers

  • HAL Id : tel-00003344, version 2

Collections

Citation

Jérémie Bigot. Recalage de signaux et analyse de variance fonctionnelle par ondelettes. Applications au domaine biomédical. Autre [q-bio.OT]. Université Joseph-Fourier - Grenoble I, 2003. Français. ⟨tel-00003344v2⟩

Share

Metrics

Record views

688

Files downloads

2459