Automated detection and classification of objective prism stellar spectra

Abstract

Abstract The spectral classification of stars in galaxies is a method of studying their stellar content. In particular, the distribution of spectral types in the stellar systems is a powerful diagnostic for the estimation of their composition, age and evolutionary state. The observational materials mainly used are spectral plates obtained with Schmidt telescopes. Such plates contain, generally, thousands of spectra. There are prism-plate libraries and digitized databases in several astronomical centres that can be fully exploited only with automated processing. Such a system for automated processing of stellar spectra on digitized objective-prism plates is presented. It is developed as a context under the MIDAS environment and provides tools for digital preprocessing of the images (rotating, correcting the tilt, calibration, etc.), for interactive and automated detection of the spectra, for spectra extraction as one-dimensional scans, and finally, for automated classification. Here we particularly present...