Translator Disclaimer
23 December 2004 Real-time implementing wavefront reconstruction for adaptive optics
Author Affiliations +
The capability of real time wave-front reconstruction is important for an adaptive optics (AO) system. The bandwidth of system and the real-time processing ability of the wave-front processor is mainly affected by the speed of calculation. The system requires enough number of subapertures and high sampling frequency to compensate atmospheric turbulence. The number of reconstruction operation is increased accordingly. Since the performance of AO system improves with the decrease of calculation latency, it is necessary to study how to increase the speed of wavefront reconstruction. There are two methods to improve the real time of the reconstruction. One is to convert the wavefront reconstruction matrix, such as by wavelet or FFT. The other is enhancing the performance of the processing element. Analysis shows that the latency cutting is performed with the cost of reconstruction precision by the former method. In this article, the latter method is adopted. From the characteristic of the wavefront reconstruction algorithm, a systolic array by FPGA is properly designed to implement real-time wavefront reconstruction. The system delay is reduced greatly by the utilization of pipeline and parallel processing. The minimum latency of reconstruction is the reconstruction calculation of one subaperture.
© (2004) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Caixia Wang, Mei Li, Chunhong Wang, Luchun Zhou, and Wenhan Jiang "Real-time implementing wavefront reconstruction for adaptive optics", Proc. SPIE 5639, Adaptive Optics and Applications III, (23 December 2004);

Back to Top