Sky-scraping boffins mash amateur astronomers into huge virtual telescope Amateur astronomers worried that Big Astronomy would render them obsolete can relax: the kinds of techniques used to create huge virtual telescopes are now being applied to the huge collections of astro-pics published on the Internet. As keen astronomy-watchers know, the effective aperture of telescopes can be expanded by …
While this technique, of necessity, can't combine the amateur telescopes into a super-telescope with the radius of the Earth for resolution - that would require them all looking at the sky at the same time, and being linked via interferometry - it is similar to the effective technique used by amateur astronomers to get rid of atmospheric blur, and get pictures of Jupiter and other planets that rival ones taken by the Hubble.
So the pictures will have better contrast, and they will look like they were taken by a telescope in space - but their resolution will still be limited by the sizes of the mirrors on the amateur telescopes they were taken with.
Well, sort of. If Intel can do photolithography of 22 nm features with light having a 195 nm wavelength, a little finagling is possible.
it is similar to the effective technique used by amateur astronomers to get rid of atmospheric blur, and get pictures of Jupiter and other planets that rival ones taken by the Hubble.
While there's obviously some borrowing of conventional image stacking techniques this isn't a simple evolution of prior art - it's perfectly justified to consider it as a novel approach. Conventional image stackers depend on an implied context that is common to all images - same target, detector, resolution, filters, intensity levels for starters. While current algorithms can compensate for target placement and image rotation artifacts you are best trying to keep those as near as possible the same too at capture since it gives greater certainty to the analysis.
The novel step here isn't the image stacking per se, it is the combination of images where none of that shared context holds and taking and combining the information present in each image even if it is not shared by the others. For example purely by eye I would suspect the image in the lower left of the example composite presented includes an IR component not present in the other images. Conventional stacking would simply discount and throw away that image since the computed Strehl ratio would be much lower than the others. However, if you look at the composite you can see that it does contribute fresh information not present in the other images.
You run the algorithm on sets of pictures on objects where you already have superior resolution pictures taken with better equipment. Then you compare, calibrate, try again until you can with confidence say that it isn't so. Kind of like what you do for any development cycle for virtually any technology.
doh!
Though I have forgotten the name of the software now, in the late '90s I bought a copy of an image enhancement program that was astonishing. It used a an algorithm based on fractal mathematics, and with an exemplar to work from, could yield some remarkable results that were at times on a par with television crime show fictional results. Running on a PC it was fairly slow, several hours in some instances, to render an image. I was pretty disappointed the next year when the company simply vanished. I have often wondered whether the whole thing was grabbed by guys in black helicopters.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
