21 July 2014 First on-sky results of a neural network based tomographic reconstructor: Carmen on Canary
Author Affiliations +
Abstract
We present on-sky results obtained with Carmen, an artificial neural network tomographic reconstructor. It was tested during two nights in July 2013 on Canary, an AO demonstrator on the William Hershel Telescope. Carmen is trained during the day on the Canary calibration bench. This training regime ensures that Carmen is entirely flexible in terms of atmospheric turbulence profile, negating any need to re-optimise the reconstructor in changing atmospheric conditions. Carmen was run in short bursts, interlaced with an optimised Learn and Apply reconstructor. We found the performance of Carmen to be approximately 5% lower than that of Learn and Apply.
© (2014) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
J. Osborn, D. Guzman, F. J. de Cos Juez, A. G. Basden, T. J. Morris, É. Gendron, T. Butterley, R. M. Myers, A. Guesalaga, F. Sanchez Lasheras, M. Gomez Victoria, M. L. Sánchez Rodríguez, D. Gratadour, G. Rousset, "First on-sky results of a neural network based tomographic reconstructor: Carmen on Canary", Proc. SPIE 9148, Adaptive Optics Systems IV, 91484M (21 July 2014); doi: 10.1117/12.2057462; https://doi.org/10.1117/12.2057462
PROCEEDINGS
6 PAGES


SHARE
Back to Top