Tomography reconstruction using the Learn and Apply algorithm

Abstract

In the framework of the MOAO demonstrator CANARY, we developped a new concept of tomography algorithm that allows to measure the tomographic reconstructor directly on-sky, using or not, a priori from the turbulence profile. This simple algorithm, working in open-loop, uses the measured covariance of slopes between all the wavefront sensors (WFS) to deduce the geometric and atmospheric parameters that are used to compute the tomographic reconstructor. This method called “Learn and Apply” (L&A) has also the advantage to measure and take into account all the misalignments between the WFSs in order to calibrate any MOAO instrument. We present the main principle of the algorithm and the last experimental results performed in MOAO scheme at the SESAME bench.