Polarized emissions from various astrophysical sources provide an additional means to infer the properties of the source. This information can be extracted by imaging the Stokes parameters characterizing these emissions, from the incomplete Fourier sampling measurements provided by the radio interferometer. To solve the corresponding ill-posed inverse problem, we recently proposed the Polarized SARA method which exploits the polarization constraint along with imposing an average sparsity prior for each of the Stokes parameters, encompassed in a reweighting approach. Motivated by the large amounts of data produced by the next-generation radio interferometers, in the current work, we propose a scalable version of this method, combined with an accelerating strategy. We showcase the performance of this improved method on a real dataset acquired by the Very Large Array (VLA).
|Title of host publication||2018 IEEE 10th Sensor Array and Multichannel Signal Processing Workshop (SAM)|
|Number of pages||5|
|Publication status||Published - 30 Aug 2018|
|Name||Sensor Array and Multichannel Signal Processing Workshop (SAM)|
Birdi, J., Repetti, A., & Wiaux, Y. (2018). Scalable Algorithm for Polarization Constrained Sparse Interferometric Stokes Imaging. In 2018 IEEE 10th Sensor Array and Multichannel Signal Processing Workshop (SAM) (pp. 465-469). (Sensor Array and Multichannel Signal Processing Workshop (SAM)). IEEE. https://doi.org/10.1109/SAM.2018.8448968