Astronomy and Astrophysics – Astronomy
Scientific paper
Dec 2005
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2005aas...207.2902d&link_type=abstract
American Astronomical Society Meeting 207, #29.02; Bulletin of the American Astronomical Society, Vol. 37, p.1211
Astronomy and Astrophysics
Astronomy
Scientific paper
In many instruments, pixel sizes are somewhat too large for the underlying optical resolution. Marginal sampling degrades the data in ways that are not obvious when one views an image. This proved to be a serious obstacle in HST/STIS/CCD spectroscopy obtained for the Eta Carinae Treasury Program -- a deficiency of available software, not of the basic data quality. In order to obtain spatial resolution that was both good and consistent, we developed an appropriate technique for "sub-pixel modeling," mathematically akin to interpolation. For reasons related to missing Fourier components, this problem is not as easy as it may appear at first sight.
Our general approach is suitable for most imaging data where the ratio of pixel width to underlying p.s.f. FWHM is between about 0.3 and 1.0. "Dithering" to improve the effective pixel size is often either impractical or unnecessary, and deconvolution is of little help for this problem. Therefore our method may be useful for a variety of astronomical data images.
The Hubble Treasury Project for Eta Carinae has been supported by STScI, which is funded by NASA.
Davidson Kris
Ishibashi Kazunori
Martin John C.
No associations
LandOfFree
Subpixel Modeling for Marginally Sampled Images does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Subpixel Modeling for Marginally Sampled Images, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Subpixel Modeling for Marginally Sampled Images will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1279293