Hassan Mortada
|
PhD Subject
Joint Decomposition using Sparse Approximation in Astronomical Multispectral Images
Funding: ANR DSIM.
Director: Christophe Collet.
Co-director: Charles Soussen.
Supervisor: Vincent Mazet.
Telescopes today routinely provide high-resolution multispectral images, i.e. 3D images where the third dimension is the wavelength : each pixel of the multispectral image is a peak (or line) spectrum. Then, the spectrum of galaxies is observed to measure the peak shift due to the Doppler effect and induced by galaxy kinematics. the goal of this thesis is to approximate the emission lines peaks in the the Astonomical images. These peaks evolve slowly between two neighbouring spectra, another challenge goal is to estimate the trajectory made by the peaks in order to decompose the image structures.
Research Interests
- Sparse approximation
- Inverse problems
- Sources separation
Talks
- Summer school Peyresq 2016: 'Modèles probabilistes et inférence en signal et image'
- GDR ISIS meeting "inverse problems", 20 mars 2017, Paris: 'Séparation de sources retardées, paramétriques et corrélées'
Teaching
2016/2017
- Signal processing (2nd year-S1)
- Image processing (Master 2 -S1)
- Numerical analysis (1st year -S1)
- 2D signal processing (2nd year -S2)
- Digital communication (2nd year -S2)
Publications
<anyweb> http://newlsiit.u-strasbg.fr/papr/appli.php?author=mortada&title=&labo=4&team=toutes&annee1=&annee2=&display=rap+&nationalRank=toutes&project=tous&hide=0&hide=0 </anyweb>