Laboratory tests of planet signal extraction in high contrast images

Understanding the formation, evolution and surprising diversity of exoplanetary system is recognized as one of the few major challenges of current astrophysics. While a large number of planets are discovered thanks to techniques like radial velocity and transits, only a few of them have clear mea- surements of their atmospheric components. Besides, these latter have been studied on transiting planets with very short orbits. Study of planets at larger separations requires direct imaging, which has enabled detection of a handful of exoplanets. This number will dramatically increase with the arrival in 2013 of SPHERE and GPI instruments that will give access to a large class of self-luminous young exoplanets. Characterization of mature planets or even massive rocky planets is expected for the next generation of planet finders that will be installed on Extremely Large Telescopes (ELT). On ELT, even with Adaptive Optics (AO) working at their best, using smart wavefront sensor and correction strategy, it is expected that the residual speckles in the images will still be a factor 100 brighter than the planet signal. This level composed of slow quasi static speckles not detected by the wavefront sensor and the rapidly varying wavefront errors that cannot be corrected by the AO loop frequency. Solutions are actually studied to calibrate these speckles and make sure that we can di erentiate them from planet signal. One of the best solution is to use the signal of focal plane wavefront sensors that can help suppressing the quasi-static speckles but also help to extract the planet signal in the final images. After describing the benefit of focal plane wavefront sensor for data extraction, we will describe our laboratory test bench which uses the Self-Coherent Camera as focal plane wavefront sensor. The principle of the data processing used to extract the planet signal will be presented together with laboratory results on very high contrast images.