论文部分内容阅读
Chinese Spectral Radio Heliograph(CSRH)is an advanced synthesis aperture solar radioheliograph,developed by National Astronomical Observatories,Chinese Academy of Sciencesindependently.It consists of 100 reflector antennas,which are grouped into two antenna arrays(CSRH-I and CSRH-II)for low and high frequency bands respectively.It is about to startofficial operation for daily observation and provided scientific data to the peer researchers.It isurgently necessary to commence imaging and data processing on CSRH.The imaging principleof CSRH lies in the well-known interferometry concerning synthesis aperture.Due to thelimited number of antennas,the captured data by CSRH is extremely sparse,and cannot recoverthe brightness of the Sun clearly.We investigate the imaging of CSRH by the aid of CompressedSensing(CS).In addition,we exploit the examples-facilitated strategy in image reconstructionof CSRH.This strategy has the following specifics:1)we take a dictionary based CS frameworkto recover signal,assuming that the signal is sparse in the given dictionary specifically designedfor CSRH,instead of using the commonly-used wavelet or Fourier transform to represent thesparsity property of signal; 2)to get a specific dictionary for CSRH,a set of good examples withhigh quality are collected to learn the dictionary,where the good examples come from availableprocessed data of CSRH and high quality observations of other solar radio observators; 3)machine learning approaches are employed to learn dictionary.