Share Email Print
cover

Proceedings Paper

Possibilities of deconvolution of image sampling structures
Author(s): Jiri Hozman
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

The paper deals with influence of the 2D sampling process upon image quality. The Optical Transfer Function (OTF), which is closely related to the Point Spread Function (PSF) of optical and electro-optical imaging systems, can be regarded as an objective measure of their quality. The main goal was the implementation of direct and blind deconvolution methods in MATLAB environment, in order to estimate these parameters and use them for computation of other characteristics, such as the Modulation Transfer Function (MTF) and the Phase Transfer Function (PTF). Relations between these functions are very useful in deriving the MTF for various geometrical shapes of elementary detectors of image sensors. This paper is focused on direct deconvolution by inverse and Wiener filtering, special focus is held on blind deconvolution using Iterative Blind Deconvolution (IBD), Simulated Annealing (SA) and Blind Deconvolution by Genetic Algorithm (BDGA). The whole process has been modeled in MATLAB. The Graphical User Interface (GUI) was also developed for setting of the deconvolution methods parameters. The parameter PSNR was also used for comparison and evaluation. The image deconvolution method based on the Blind Deconvolution by Genetic Algorithm seems to be very useful, especially from the point of view of computational requirement and results as well.

Paper Details

Date Published: 18 April 2006
PDF: 6 pages
Proc. SPIE 6180, Photonics, Devices, and Systems III, 61801X (18 April 2006); doi: 10.1117/12.675836
Show Author Affiliations
Jiri Hozman, Czech Technical Univ. in Prague (Czech Republic)


Published in SPIE Proceedings Vol. 6180:
Photonics, Devices, and Systems III
Pavel Tománek; Miroslav Hrabovský; Miroslav Miler; Dagmar Senderákova, Editor(s)

© SPIE. Terms of Use
Back to Top