- "Er, what?" I hear you mutter. OK, here's what it might
be able to do:
Centre of a large cluster of galaxies observed with different spatial resolutions. When spatial resolution improves, finer details can be distinguished (as, in this image, the arcs formed by gravitational lensing).
The technique is used in microscopy and, as you can see from the image above, in astrophotography. There are a number of commercial and free packages out there which can apply deconvolution but the "MCS" algorithm used in that image isn't one of them, so far as I know.
Maybe one day one may be able to buy it as, from page 15 of this paper
, they are getting to the point where a graphics processor can be used to accelerate processing of images 2048 pixels on a side in about an hour on a laptop. The market may be small but the benefits might be huge for those of us without access to full-on adaptive optics replete with high powered lasers and mountain-top observatories.
As a final example check out these deconvolutions of a simulated image of a star cluster partly superimposed on a background galaxy.
Top left: true light distribution with 2 pixels FWHM resolution;
Bottom left: observed image with 6 pixels FWHM and noise;
Top middle: Wiener filter deconvolution of the observed image;
Bottom middle: 50 iterations of the accelerated Richardson-Lucy algorithm;
Top right: maximum entropy deconvolution;
Bottom right: deconvolution with the MCS algorithm.
Yep, bottom left is what you might get from your camera and bottom right is MCS Deconvolution of that image, it's far and away the best in this simulated run as one can see by comparing it with the original image at top left and you can't buy it.
Any academics out there want to take on the challenge of making this algorithm more widely available? I really hope so.
P.S. Although this is an "image processing" thread I think it belongs in this section because it is only of relevance to "Technical and scientific photography".
Upon revisiting this thread I found a number of links and images were no longer available so I've updated the post as well as I could.