2008MNRAS.385..918K

We present a semi-automated method to search for strong galaxy-galaxy lenses in optical imaging surveys. Our search technique constrains the shape of strongly lensed galaxies (or arcs) in a multiparameter space, which includes the third-order (octopole) moments of objects. This method is applied to the Deep Lens Survey (DLS), a deep ground-based weak lensing survey imaging to R ∼ 26. The parameter space of arcs in the DLS is simulated using real galaxies extracted from deep Hubble Space Telescope fields in order to more accurately reproduce the properties of arcs. Arcs are detected in the DLS using a pixel thresholding method and candidate arcs are selected within this multiparameter space. Examples of strong galaxy-galaxy lens candidates discovered in the DLS F2 field (4deg²) are presented.