A highly efficient lucky imaging algorithm: Image synthesis based on fourier amplitude selection

We propose a new algorithm dramatically enhancing the efficiency of the lucky imaging technique for AO-corrected images in the visible range. It is achieved by a selection based on the relative strength of signal for each spatial frequency in the Fourier domain, making a more efficient use of information contained in each frame. Realistic simulations show that our algorithm allows us to reach the diffraction limit in the visible range on an AO-equipped 8 m telescope and enhances the Strehl ratio of an AO long exposure by a factor of up to 4. It outperforms the lucky imaging technique at an equivalent selection ratio. The fraction of selected data in simulation is also boosted from two to eight times for a given Strehl-ratio performance.