Multiscale and multidirectional very long baseline interferometry imaging with CLEAN

Context. Very long baseline interferometry (VLBI) is a radio-astronomical technique whereby the correlated signal from various baselines combined into an image of highest possible angular resolution. Due to sparsity measurements, this imaging procedure constitutes ill-posed inverse problem. For decades, CLEAN algorithm has been standard choice in VLBI studies, despite it bringing on some serious disadvantages and pathologies that are brought by requirements modern frontline applications. Aims. We developed novel multiscale deconvolution method (DoB-CLEAN) based continuous wavelet transforms address several imaging. benchmarked against reconstructions synthetic data reanalyzed BL Lac observations RadioAstron with DoB-CLEAN. Methods. The DoB-CLEAN approaches via multiscalar multidirectional dictionaries. Two different dictionaries were used: 1) difference elliptical spherical Bessel functions dictionary fitted uv -coverage observation used sparsely represent features dirty image; 2) Gaussian well suited relevant cleanly. was performed switching between Results. achieves super-resolution compared remedies spurious regularization properties CLEAN. In contrast CLEAN, representation basis physical meaning. Hence, computed deconvolved still fits observed visibilities, Conclusions. State-of-the-art seem outperform single-scalar maximize extraction information ongoing