Filament and Flare Detection in Hα image sequences
This work addresses solar storm monitoring for solar physicists, enabling near real-time alerts and statistical analysis, but appears incremental as it builds on existing segmentation and classification approaches.
The paper tackles the problem of detecting solar flares and filaments in Hα image sequences by proposing a new method involving preprocessing, variational segmentation, and postprocessing classification, demonstrating performance through comparison with expert annotations.
Solar storms can have a major impact on the infrastructure of the earth. Some of the causing events are observable from ground in the Hα spectral line. In this paper we propose a new method for the simultaneous detection of flares and filaments in Hα image sequences. Therefore we perform several preprocessing steps to enhance and normalize the images. Based on the intensity values we segment the image by a variational approach. In a final postprecessing step we derive essential properties to classify the events and further demonstrate the performance by comparing our obtained results to the data annotated by an expert. The information produced by our method can be used for near real-time alerts and the statistical analysis of existing data by solar physicists.