Previous abstract Next abstract
The IUE Final Archive data reduction system, NEW Spectral Image Processing System (NEWSIPS), was specifically designed to process the entire IUE dataset in a consistent and homogenous fashion. The system is comprised of enhanced image processing techniques which exploit the inherent characteristics of IUE data in order to achieve ultimately an improved signal-to-noise ratio (S/N) in the extracted spectrum.
All raw IUE images suffer from: non-linearities due to detector response, fixed pattern noise, periodic noise, and multiple geometric distortions. In addition, the dispersion direction lies at an angle with respect to either image axis, rendering spectral extraction complicated. NEWSIPS was designed to address each of these challenges by applying rigorous mathematical and image processing techniques.
NEWSIPS includes: automated raw image characterization; Intensity Transfer Functions (ITFs) created in their own ``geometric'' space -- eliminating the need for resampling (and consequent degradation) of calibration data; an explicit image registration combined with the nondegraded ITFs yield a more accurate photometric correction; a single flux- and line-shape preserving resampling to a rectified spatial domain which incorporates multiple ``geometric corrections'' such that the spectral orders are parallel to an image axis and the wavelength dispersion is linear within each order; and a signal-weighted extraction method which includes an error estimate for each extraction bin.
NEWSIPS achieves a significant increase in the S/N in the two-dimensional images which ultimately manifests itself in the extracted spectra. The increase in the S/N ranges from 10 -- 200\%; the greatest improvements occurring in under-exposed, high radiation, and high sky background images. Examples of the improvements in the low-dispersion data are presented.
This work was supported under NASA contract NAS5-31230 to CSC.
Thursday program listing