Automatic Identification of Solar Granules and Magnetic Bright PointsAutomatic Identification of Solar Granules and Magnetic Bright Points

Safari Hossein, safari@znu.ac.ir, Department of Physics, University of Zanjan, Iran
Javaherian Mohsen, m_javaherian@znu.ac.ir, Department of Physics, University of Zanjan

Abstract
In this study, based on automatic approach, solar granules (Gs) and magnetic bright points (MBPs) are identified. To do this, the images at two different wavelengths (214 and 313 nm) recorded by Hinode are used. The features are classified by supporting vector machine (SVM) classifier. With removing the MBPs pixels from original image, the granular pixels are segmented based on 2-D wavelet transformation. The physical properties (size distributions, etc) of both Gs and MBPs are calculated for a large number of images.