Statistics – Computation
Scientific paper
Jul 1993
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=1993phdt.........2m&link_type=abstract
Ph.D. Thesis Queens Univ., Kingston (Ontario). Dept. of Electrical Engineering.
Statistics
Computation
1
Astronomical Photography, Atmospheric Correction, Image Reconstruction, Speckle Interferometry, Algorithms, Digital Simulation, Least Squares Method, Mathematical Models
Scientific paper
A new quasi least-squares (QLS) algorithm has been developed for phase recovery in the triple correlation (TC) process applied to speckle imaging in astronomy. The QLS algorithm is found to provide significant improvement in terms of image reconstruction fidelity over the popular recursive algorithm. Implementing the QLS algorithm using phasors avoids the modulo 2-pi problem inherent to linear least squares solutions. The performance of the QLS algorithm, along with a Knox-Thompson (KT) phase recovery process for comparison, was rigorously studied using second- and third-order noise models. Those models eliminate the need for speckle image simulation and the computation of second- and third-order products required by KT and TC for phase recovery. The validity of the model was confirmed using simulated speckle data. Simulations showed that for the high signal to noise ratio (SNR) case, the near-axis TC transfer function subplanes have lower phase error than the comparable KT transfer function subplanes. Computing subplanes only in the near-axis region is important if computer resources are limited. Because of the lower phase error, the TC process was found to produce better images than the KT algorithm with high-SNR input data. This previously unreported difference is analyzed in detail.
No associations
LandOfFree
Properties of the near-axis bispectrum applied to high-resolution astronomical imaging does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Properties of the near-axis bispectrum applied to high-resolution astronomical imaging, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Properties of the near-axis bispectrum applied to high-resolution astronomical imaging will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1074238