Open AccessScale sequentially CLEAN for Mingantu Spectral Radioheliograph
Author(s) -
Чэн Цзюнь,
Jun Cheng,
Йан Йихуа,
Yihua Yan,
Чжao Дун,
Zhao Dong,
Сюй Лун,
Long Xu
Publication year - 2019
Publication title -
solnečno-zemnaâ fizika
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.11
H-Index - 2
ISSN - 2412-4737
DOI - 10.12737/szf-52201908
Subject(s) - wavelet , deconvolution , computer science , aperture (computer memory) , optics , scale (ratio) , image resolution , aperture synthesis , sampling (signal processing) , remote sensing , computer vision , physics , acoustics , geology , filter (signal processing) , quantum mechanics , interferometry
MingantU SpEctral Radioheliograph (MUSER) is a solar-dedicated radio heliograph, adopting aperture synthesis technique to image the Sun in the frequency range of 0.4 GHz to 15 GHz. MUSER has extremely high spatial resolution, temporal resolution, and frequency resolution beyond those of contemporary devices of the same category. For aperture synthesis, the number of antennas is limited, so sparse sampling of Fourier components is actually obtained for solar observation, which corresponds to the situation that a clean image is convolved by a dirty beam with strong sidelobe in a spatial domain. Thus, the deconvolution, such as CLEAN, is generally required for imaging the aperture synthesis to remove artifacts caused by the convolving dirty beam. The traditional Högbom CLEAN is based on the assumption that an observed object is only composed of point sources. This assumption does not hold for solar observation, where the solar disk is an extended source containing complex structures and diffuse features. In this paper, we make the first attempt to employ scale sequentially CLEAN for MUSER imaging, including Multi-Resolution CLEAN and Wavelet CLEAN. The experimental results demonstrate that the scale sequentially CLEAN, especially wavelet CLEAN, is superior to the traditional CLEAN algorithm in smaller number of iterations and improved image quality. We provide optimized wavelet parameters to further improve the performance of wavelet CLEAN.