You are here: Home > Teaching



Inverse Problems for Image Reconstruction

These courses introduce inverse problem approach using examples based on image reconstruction. Below you can find different versions:

É. Thiébaut, "Introduction to Image Reconstruction and Inverse Problems," Optics in Astrophysics, Springer, Foy, R. & Foy, F.-C. (Eds.) 198, 397-422 (2005, text in English).

É. Thiébaut & Ch. Pichon, "Approche bayesienne pour la résolution de problèmes inverses,"  VIIIème Ecole de Cosmologie, Cargèse, 28 août – 2 septembre 2006 (slides in French).

"Approche inverse en traitement du signal," transparents du cours, 2017 (slides in French)

Image deconvolution is the subject of the proposed practical sessions.  In order to deal with real data, images from the HST (Hubble Space Telescope) can be found at STSI FTP Archive for Saturn, Jupiter and the super-nova SN1987a. The exercices can be done with any data processing language such as Yorick, IDL or MatLab.

- Instructions for the practical session (in French) tp-Lyon-2012-instructions.pdf.

- Files for practical session on image deconvolution:
    - scripts for Yorick;
    - scripts for MatLab;

- Yorick interpreter for data processing practical session:

-Quich instructions for Yorick (in French) yorick-rapide.pdf.

- MS-Windows (32-bit version):

1. get zip archive here
2. unzip in some directory
3. launch interpreter by double clicking of yorick.exe in bin subdirectory
4. change to your working directory: > cd, "~/work";

- Unix-like version (Linux/Darwin):

1. download archive here (.tar.bzip2, 21 Mb) or here (.tar.xz, 8.1 Mb)
2. extract archive contents:
  tar jxvf easy-yorick-VERSION.tar.bz2
  tar Jxvf easy-yorick-VERSION.txz
3. start executable in easy-yorick-VERSION/bin/yorick

To edit the Yorick script, you'll need a descent text editor, under Linux gedit, emacs or vi are perfect, under Windows, you may try notepad++.

Image Reconstruction for Optical Interferometry

Thesis supervision

Yeti 6.4.0

Yeti version 6.4.0 has been released (2015-02-06). This version has no new feature compared to 6.3.3 but is now hosted on GitHub. [more]


MiRA won 2008' Beauty Contest

MiRA has won the 2008' edition of the Beauty Contest for image reconstruction algorithms from optical interferometry data.