Fast automated detection of CMEs fronts relevant for the Space Weather Program

In this paper we present a new algorithm to perform near real-time automatic detection of Coronal Mass Ejections (CMEs) with STEREO (Solar Terrestrial Relations Observatory) COR-2 beacon images (256°256 pixels). The aim of the algorithm is to generate trustable data for the Space Weather Program (SWP) using image processing techniques, thus replacing the actual manual extraction methods. We present the results of the use of the algorithm on the October 10, 2010 and the April 8, 2010 CMEs. We extract values of distance, speed, width and direction frame by frame for each CME detected by the algorithm as no user intervention is required. The results are compared with manually extracted boundaries from the SWP with the DICE coefficient and other comparison indicators, reflecting the good performance of quality and time. These results show the detection of 100% of relevant CMEs, processing each frame in less than 200 ms and generating DICE coefficients above 80% when compared to manual extraction. We explain those results and present an application to visualize the results of the algorithm and to compare it with other boundaries: the Cme Detector. This application will be released to a small group of solar astrophycisists by february 2011.