NVST chromosphere data interference fringes removal based on NSCT and PCA

Abstract The New Vacuum Solar Telescope (NVST) has been affected by thin film interference since its operation, which makes interference fringes superimposed on its sun chromosphere data products. These fringes have a serious impact on the quantitative analysis and high-resolution reconstruction of solar fine structure observation. Although some methods have been adopted to remove the interference fringes in NVST chromosphere data, most of them are performed on high-resolution reconstructed data, and little for short-exposure data (speckle images). The short-exposure data are the raw data of the NVST observation. If the fringes on the data can be eliminated or weakened, the quality of subsequent high-resolution reconstructed data will be improved. Therefore, this paper proposed a method to remove the interference fringes in short-exposure data based on Non-Subsampled Contourlet Transform (NSCT) and Principal Component Analysis (PCA). First, NSCT multi-scale decomposition is utilized to divide the data into fringe-containing scales and fringe-free scales, and then the fringes are removed by PCA in fringe-containing scales. Experiments show that the interference fringes in the data can be removed effectively while preserving the solar structures of NVST data.