The Light Curve Statistical Moments Analysis: The Identification of Eclipsing Binaries

A semi-automatic method of identifying eclipsing binaries in large photometric surveys, based on the analysis of the light curve statistical moments, is presented. This method was tested on the OGLE-II database of variable stars in the Large Magellanic Cloud. As a result 99.0% of eclipsing binaries in Wyrzykowski et al. (2003) catalog and 100% of eclipsing binaries in Groenewegen (2005) catalog were recovered. Additionally we identified 752 new eclipsing binaries in the LMC that were not presented in earlier catalogs thereby increasing detection rates by about 25%. The method is more focusing on detached and semi-detached eclipsing binaries. We conclude that this method is well suited for use in eclipsing binary searches of large surveys such as OGLE-Ill.