Offre en lien avec l’Action/le Réseau : – — –/– — –
Laboratoire/Entreprise : Institut de planétologie et d’Astrophysique de Gr
Durée : 4-6 months
Contact : julien.milli@univ-grenoble-alpes.fr
Date limite de publication : 2021-02-15
Contexte :
The VLT/SPHERE high-contrast imager is one of the most powerful instruments for the detection of exoplanets and discs by direct imaging, combining extreme adaptive optics and coronagraphy. The main challenge in data processing consists in estimating and then subtracting as much as possible the light from the central star that has not been entirely blocked by the coronagraph, to reveal in the circumstellar environment the potential signature of exoplanets in orbit around the star, or dust rings leftover from the planetary system formation. Currently, the most efficient techniques are based on angular diversity, which allows to estimate stellar residuals empirically from a temporal sequence of images of the target star where the telescope pupil is fixed and objects in the sky rotate. The main difficulty of this approach is to properly separate the stellar residuals from the astrophysical signal of planets or disks during estimation. This is especially true for extended signals from discs, making this technique almost blind to discs seen under a pole-on configuration.
An alternative technique is to use a library of images of other stars to estimate the stellar residuals. Thanks to recent developments in algorithms and advances in computing capacity, this technique is now attracting great interest in ground-based instruments fed by extreme adaptive optics, such as VLT/SPHERE, which compensate for the atmospheric turbulence.
Sujet :
The aim of the internship is to improve the image processing of the VLT/SPHERE instrument with the help of image libraries. In particular, the study will focus on the impact of the size of the library on the performance. This will be done using a dataset of 26 target stars, allowing the performance to be characterised for an average library size of about 4000 images. This will be compared with the results obtained by compiling a library one hundred times larger, using the architecture of the SPHERE Data Center in Grenoble. Particular interest will be paid to the performance of extended signals detection, such as discs, which is one of the major interests of this technique.
This study is part of and funded by the ERC project COBREX, which brings together researchers from the Laboratoire d’Etudes Spatiales et d’Instrumentation en Astrophysique (LESIA, at the Paris Observatory), the Institut de Planétologie et d’Astrophysique de Grenoble (IPAG) where the internship will take place, and the Centre de Recherche Astrophysique de Lyon (CRAL).
Profil du candidat :
The candidate should have skills either in the following areas: signal and image processing, machine learning, optics
Formation et compétences requises :
The candidate can have a background either in applied mathematics/ data science or in astrophysics / instrumentation for astrophysics
Adresse d’emploi :
Institut de Planetologie et d’Astrophysique de Grenoble (IPAG)
414, rue de la Piscine
Domaine Universitaire Saint-Martin-d’Hères
BP 53
38041 Grenoble cedex 9
France