Abstract

For several convolution equations such as f = g*u, the support domain of the unknown being bounded, optical resolution becomes difficult as soon as f does not decrease quickly enough at infinity. When the approximation g(x − y) ≈ g(x) is good for |x| ≫ |y|, we are able to correct by interferometric means the necessary truncation of the f information. In particular, the proposed correction is interesting for the optical treatment of the primary data used in transaxial tomography.

© 1979 Optical Society of America

Convolution algorithm for the Lorentz function

Shepard A. Clough and Francis X. Kneizys
Appl. Opt. 18(13) 2329-2333 (1979)

Halftone screen techniques for photographic film response shaping with application to optical processing of medical x-ray images

I. Glaser and H. H. Barrett
Appl. Opt. 18(13) 2294-2300 (1979)

Reconstruction of a Fraunhofer hologram recorded near the focal plane of a transforming lens

K. Chałasińska-Macukow and T. Szoplik
Appl. Opt. 18(9) 1436-1441 (1979)

Noise performance of an achromatic coherent optical system

Emmett N. Leith and James A. Roth
Appl. Opt. 18(16) 2803-2811 (1979)

Optical transfer function synthesis: a geometrical optics approach

H. H. Barrett, M. Y. Chiu, S. K. Gordon, R. E. Parks, and W. Swindell
Appl. Opt. 18(16) 2760-2766 (1979)