An Automated Method of Selecting Galactic Stars with DIB Features from the SDSS

The problem of Diffuse Interstellar Bands (DIBs) has been studied for almost a century since the discovery, but it still remains a mystery. A method is proposed in this paper for automatically selecting Galactic stars with DIB features from the SDSS DR7 data. This method is based on spectral-feature matching, and selecting of DIB features by requiring sufficiently low relative neighboring-pixel fluxes for the central-wavelength pixels of DIB lines (-line-profile prominence). We are able to select candidates of Galactic stars with DIB features with this method from calibrated spectra, as shown by experiments on over 300 SDSS plates. The paper thus shows the simplicity, effectiveness, and robustness of the method. The method will provide a huge amount of data for research areas such as identifications of stars of these features, and will push forward research of DIBs.