Automated sunspot detection using morphological reconstruction and adaptive region growing techniques

Sunspots are dark areas with respect to their surrounding area because the temperature of sunspot areas is lower than the average temperature of the solar surface. It provides essential information for many aspects of solar physics. A sunspot is composed of umbra and penumbra. It is a prerequisite for studying solar physics and spatial atmosphere to accurately segment and extract sunspot structures. Consequently, we propose an automated detection technique for segmenting and extracting them. The detection procedure is composed of two steps: (1) segmenting and extracting sunspot umbra with morphological reconstruction; (2) detecting and segmenting sunspot penumbra with region growing. For evaluating the accuracy of the detection procedure, we used a high-resolution observation obtained by Solar Optical Telescope onboard Hinode, and other obtained with the Dutch Open Telescope to illustrate the performance. The results demonstrate that our proposed technique is significant effective and accurate, and is suitable for studying the sunspot evolution and their physical phenomena.