| Popis: |
We introduced the use of Artificial Neural Networks (ANN) for centroiding in Shack-Hartmann wavefront sensors in the presence of elongated spots, as it will occur in Extremely Large Telescopes. We showed in simulation that ANNs can outperform existing techniques, such as the Matched Filter. The main advantage of our technique is its ability to cope with changing conditions, as real atmospheric turbulence behaves. Here we present experimental results from the laboratory that confirm the findings in our original article, while at the same time they are useful to refine the ANN-based techniques. |
