COMPARISON OF IMAGE-PROCESSING METHODS TO EXTRACT SOLAR FEATURES

摘要

Large numbers of high precision solar images are now available from both terrestrial and space observatories, which has made it necessary to develop automated image processing techniques. In this paper we compare analysis of two sets of full-disk solar images: ground-based white light photographic films from Gyula and allied observatories, and magnetograms and quasi-continuum images from the Michelson Doppler Imager (MDI) on SOHO. We use two different automated image analysis techniques. The Sunspot Automatic Measurement (SAM) program has been developed at the Heliophysical Observatory at Debrecen for compiling the Debrecen Photohe-liographic Data which are used to measure and catalogue the area and position of sunspots (umbra and penumbra). This project is part of the continuation of the Greenwich Catalogue. Startool is a general image analysis tool developed at the Jet Propulsion Laboratory, and applied to the MDI imagery as part of the SOHO Guest Investigator Program. As used with MDI, Startool identifies sunspots, faculae/network, and quiet sun using statistical pattern recognition techniques. Here we compare the area of sunspots as derived by Startool from the MDI images and by the SAM program as derived from the Debrecen and MDI images for the pilot interval of the second half of the year 1996.