Automated Identification of Sunspot Penumbral Grains Using Morphological Reconstruction Technique

A sunspot usually consists of a relatively dark region, called umbra, and a region surrounding the umbra, called penumbra. The penumbra includes various fine structures, like bright filaments and dark fibrils. Small and bright penumbral grains locate on the bright filaments. Studying the physical properties of penumbral grains can help us to better understand magnetic field motions in a penumbra and its formation mechanism. So we presented an automated identifying method for extracting the grains accurately. The identification method about the grains of penumbrae is composed of four steps. (1) Original image is smoothed as a mask image. Than smoothed image is eroded that as a marker image. Subsequently, the mask image is reconstructed by the marker image. (2) Many bright features are found in the difference image which is obtained between the smoothed image and reconstructed image. (3) A binary image of the difference image is obtained by adjusting the threshold value. (4) The small area bright points in the binary image are removed. In order to evaluate the accuracy of this method, we used three high-resolution images obtained with the New Solar Vacuum Telescope, the New Solar Telescope, and the Solar Optical Telescope to illustrate the performance.