Research On The Solar Brightness Model And Its Application Based On Statistics

High-quality image output is an important part of astronomical research.In the imaging process of MingantU SpEctral Radioheliograph(MUSER),the deviation of the solar disk from the center of field of view results in the poor quality of final imaging,the absence of statistical information from the original dirty image in the process of dirty image cleaning,which results in a large amount of iteration time overhead and the data processing system is not perfect because of the lack of checking after data elimination.This paper focus on a statistical-based solar brightness model,to calculate the possible brightness values of solar disk and sky background in MUSER original dirty image more efficiently.These two values and the model are applied to subsequent processing,which has played a certain application value.The main research results are as follows:(1)Analyzing the research status of the possible brightness of solar disk and sky background at home and abroad,combining the current methods of image segmentation,data fitting,anomaly detection and elimination,focus on a statistical-based solar brightness model.Through simulation test and MUSER observation data test,it is proved that the model can effectively achieve multi-Gaussian fitting,and can analyze the possible brightness values of solar disk and sky background from MUSER original dirty image.Compared with traditional GMM,the model has more stable performance.(2)Studying the reason why the solar disk deviates from the center of field of view in the MUSER original dirty image,focus on an algorithm for correcting the solar disk.The possible brightness value of the solar disk is used as the threshold value of cleaning,and the offset in the spatial frequency domain is calculated by correlation coefficient calculation,so as to improve the quality of imaging.(3)Studying the CLEAN algorithm processing flow and cleaning effect in MUSER cleaning process,taking the possible brightness value of sky background as the threshold value of CLEAN algorithm,achieving the desired cleaning purpose,speeding up cleaning speed and saving iteration time.(4)Studying the status of MUSER distortion data labeling and eliminating,analyzing the variation rule of solar brightness model when there is distortion data,and inferring whether there is distortion data in visibility data.The above research contents are tested with the data of MUSER since the trial observation from 2014.The results show that the statistical-based solar brightness model can achieve multi-Gaussian stable fitting.After analyzing the possible brightness values of the solar disk and sky background in the MUSER original dirty image,it can provide threshold reference in different processing cleaning.The model can also play an important role in the detection of outlier data rejection.