Filament recognition and image cleaning on Meudon Hα spectroheliograms

This paper presents the techniques developed for the automated detection of filaments on Meudon H alpha spectroheliograms, and, by extension, on any full-disk H alpha Sun observations. Some cleaning processes are first applied to the images to correct them from defects characteristic of the instrument. Indeed, these defects may lead to spurious detections. From the cleaned images, filament areas are then segmented using a region growing method which efficiently returns the full extent of these dark areas. The filaments are finally described by means of their pruned skeleton. This representation allows one to compare the automatically segmented filaments with those manually recorded for Meudon Synoptic Maps. The very good agreement observed on a representative set of images confirms that this method can effectively be used in the frame of the EGSO (European Grid of Solar Observations) project in order to produce a reliable catalog dedicated to solar features.