Abstract
We introduce torchmfbd, an open-source Python software package designed for Multi-Object, Multi-Frame Blind Deconvolution (MOMFBD) to restore astronomical images affected by atmospheric turbulence. Built on the PyTorch library, the code leverages automatic differentiation and GPU acceleration, significantly reducing computation time compared to traditional CPU implementations. It provides a highly flexible, modular interface that allows users to easily add observations, define phase diversity channels, and incorporate various regularization techniques (such as smoothness and sparsity constraints) to stabilize solutions and handle spatially variant point spread functions (PSFs).
Figure 1: Basis functions of the NMF dictionary for the CRISP instrument at 8542 Å. The first 100 basis functions, out of a total of 150, are shown.
