A sharpness criterion for space-variant imaging through a random medium was recently proposed. It uses ensemble-averaged images to provide the information ordinarily given by space invariance. This criterion is used to construct an algorithm for the correction of randomly distorted one-dimensional images degraded by Gaussian-distributed stretching functions of variable correlation length. It is shown theoretically and demonstrated in the simulation that knowledge of the shape of the distortion-induced transfer function allows the width of the function to be determined from the processing procedure. In the simulation, distortion-degraded sinusoids with an average visibility of 2% before processing yield a resultant sinusoid with a visibility of 80–90% after processing. Under equivalent conditions, corrections of block objects may be carried out with 100% accuracy. It should be possible to use a similar procedure to improve real imagery degraded by atmospheric turbulence.
© 1988 Optical Society of AmericaFull Article | PDF Article
OSA Recommended Articles
J. Opt. Soc. Am. 71(10) 1172-1175 (1981)
G. R. Ayers, M. J. Northcott, and J. C. Dainty
J. Opt. Soc. Am. A 5(7) 963-985 (1988)
Brian J. Thelen, Richard G. Paxman, David A. Carrara, and John H. Seldin
J. Opt. Soc. Am. A 26(1) 206-218 (2009)