During my final year internship in engineering, I had to develop a new method to reconstruct blurred astronomical images. The method is based on deconvolution using both sparsity and image moments constraints. On a wider scale, this work may be of use to weak gravitational lensing missions such as Euclid and help unveil some of the mysteries of the Universe.
Recovery of radio images
Currently, I am a PhD student and the title of my thesis is Sparse semi-parametric recovery of astronomical radio-images. The challenge is to be able to process very large images of 200 megapixels for example, and to take at the same into account Direction Dependent Effects. The goal is to propose a new framework for analyzing radio data set, where restoration and interpretation steps are done jointly. This involves a semi-parametric approach where regularization is performed in the parameters space.