Signal processing methods for high-contrast observations of planetary systems

Direct imaging of planetary systems offers valuable astrophysical information but remains a challenging task, mainly due to the tiny angular separation between faint circumstellar signal and the overwhelming bright star. Even with state-of-the art instruments, this task heavily relies on post-processing. Indeed, the acquired images are plagued with bright quasi-static spots, called speckles, that hide the astrophysical signal. These speckles prevent planetary imaging from a single image. Planetary systems imaging is made possible by dedicated observation strategies developed to add diversity into the data. A popular observation strategy is angular differential imaging (ADI) that relies on the fact that speckles remain quasi-static during a night of observation while the circumstellar signal follows a circular motion on the field of view induced by the rotation of the Earth. Post-processing techniques of ADI datasets aim to separate the signal into a temporal static component for the speckle field and a rotating component for the circumstellar signal. In the first part of this thesis, we propose signal processing methods for exoplanets detection while the second part focuses on circumstellar disks imaging. The methods presented in this thesis have been validated on both synthetic and real datasets. In the latter, we are able to extract astrophysical signal with an unprecedented precision, leading to a better understanding the nature of planetary systems and their formation process.