2012A&A...546A..10L

In the exoplanetary era, the Kepler spacecraft is causing a revolution by discovering thousands of new planet candidates. However, a follow-up program is needed to reject false candidates and fully characterize the bona-fide exoplanets. Our main aims are to 1./ detect and analyze close companions inside the typical Kepler point spread function (PSF) to study whether they are the responsible for the dimming found in Kepler light curves, 2./ study the change in the stellar and planetary parameters caused by an unresolved object, 3./ help validate the Kepler objects of interest (KOI) that do not have any object inside the Kepler PSF, and 4./ study the multiplicity rate of planet-host candidates. Such a large sample of observed planet-host candidates allows us to derive statistics for close (visual or bounded) companions to the harboring star. We present lucky imaging observations for a total of 98 KOIs. This technique is based on the acquisition of thousands of very-short-exposure-time images. A selection and combination of a small amount of the highest quality frames provides a high resolution image with objects having a 0.1-arcsec PSF. We apply this technique to carry out observations in the Sloan i and z filters of our Kepler candidates. We find blended objects inside the Kepler PSF for a significant percentage of KOIs. On the one hand, only 58.2% of the hosts do not have any object within 6-arcsec. On the other hand, we find 19 companions closer than 3 arcsec in 17 KOIs. According to their magnitudes and i-z colors, 8 of them could be physically bound to the host star.