de     

simg wiener

Deconvolves an image using the Wiener method. This method is good for images with a low dynamic range, e.g. images of planets or terrestrial images.

Usage

usage: /usr/local/bin/simg wiener [<options>] [@<parameter file>] [<infile>] [<outfile>]

parameters:
  <infile>   Source image (default: stdin)
  <outfile>  Destination image (default: stdout)

options:
  -s <nn.n>    Radius of the Gauss-shaped PSF(default: 1.5)
  -psf <file>  PSF (instead Gauss-shaped PSF)
  -snr <nn.n>  Mean SNR (default: 10)
  -o <outfile> Destination image
  -h           Help

Mathematical description

Graph

Graph

Description

This method works good for images with a low dynamic range, e.g. images of planets or terrestrial images, even if no exact PSF is known. If the PSF is given it can be specified with the -psf option, else a Gauss-shaped PSF with the radius defined by -s is generated. The option -snr is used for the calculation of the regularization parameter λ.

This deconvolution method is fast and easy to use. If no exact PSF is known proper values for the options -s and -snr can determined as follows.

  1. Start with a not too high SNR (e.g. 5 or 10) and a small PSF radius s (e.g. 0.8)
  2. Increase s until artifacts are visible in the output image
  3. Decrease s by approx. 10%
  4. Increase the SNR parameter until the desired compromise of sharpness and noise is reached

A similar deconvolution method, but with local damping, is the Damping regularization method.

 
ref/simg-wiener.txt · Last modified: 2010/03/08 09:51 (external edit)
 
Recent changes RSS feed Creative Commons License Powered by PHP Debian Driven by DokuWiki