The IRAS survey array produced sky brightness measurements with higher spatial resolution in the in-scan direction than in the cross-scan direction. To reduce processing requirements, the first step in the data reduction to produce the ISSA images was to smooth and resample, at two samples per second, the time-ordered detector data streams to an in-scan resolution of 3.5', 3.5', 3.6' and 4.7' FWHM at 12, 25, 60 and 100 µm, respectively (Appendix B). Time delays were introduced independently for each detector in the smoothing and resampling process in order to rephase the data streams to sample simultaneously the same cross-scan line on the sky.
The processing removed a zodiacal emission model from the smoothed and rephased data, refined the detector zero point stability with two destriping algorithms and binned the data into images. Separate images of the individual HCONs were produced for each field and were visually examined to allow identification and removal of artifacts (§III.D.3). After removal of artifacts by editing the time-ordered data, new images were produced for each HCON and all HCONs were then co-added. Known asteroids remain in the individual HCON images but were removed from the data prior to producing the co-added images. Auxiliary images of sky coverage and statistical noise were also produced for each ISSA field. Due to volume constraints and limited utility (§III.C.4), the sky coverage and statistical noise images are not included in the released set of images. They are, however, available upon request at IPAC.
The co-added images contain some remaining features that do not confirm among the individual HCONs. These features include fine structure in the zodiacal cloud (the zodiacal dust bands, for instance), planets, unknown asteroids and orbital debris that escaped the artifact removal process. The individual HCON images enable users to identify these nonconfirming objects, both to avoid confusion with confirming sources and to study the nonconfirming sources.