A Survey of Solar Filament Detection Methods

Solar filament is one of the main features of the sun. As a forecasting aid for geomagnetic storms, also, it is the subject of space weather study. Many methods have been proposed to detect the solar filaments, but most of them are manual feature recognition techniques. In this paper, three kinds of automated detection methods for filaments are presented. They are respectively automated region growing procedure, a threshold and region growing combined method and Artificial Neural Network technique (ANN). The experimental results are presented and the advantages and disadvantages of these three methods are compared. Automated region growing method can represent more details of filaments, however, it is time-consuming and strongly dependents on the background cleaning procedures. Threshold and region growing combined method can effectively improve the speed of image processing, but it can't detect the filaments near the solar limb. ANN can resolve these drawbacks which exist in automated region growing procedure and it has the ability of recognizing multiplicative filaments.