The development of sensitive large-format imaging arrays for the infrared p
romises to provide revolutionary capabilities for space astronomy. For exam
ple, the Infrared Array Camera (IRAC) on SIRTF will use four 256 x 256 arra
ys to provide background limited high spatial resolution images of the sky
in the 3 to 8 mu m spectral region. In order to reach the performance limit
s possible with this generation of sensitive detectors, calibration procedu
res must be developed so that uncertainties in detector calibration will al
ways be dominated by photon statistics from the dark sky as a major system
noise source. In the near-infrared, where the faint extragalactic sky is ob
served through the scattered and re-emitted zodiacal light from our solar s
ystem, calibration is particularly important. Faint sources must be detecte
d on this brighter local foreground. We present a procedure for calibrating
imaging systems and analyzing such data. In our approach, by proper choice
of observing strategy, information about detector parameters is encoded in
the sky measurements. Proper analysis allows us to simultaneously solve fo
r sky brightness and detector parameters and provides accurate formal error
estimates. This approach allows us to extract the calibration from the obs
ervations themselves; little or no additional information is necessary to a
llow full interpretation of the data. Further, this approach allows refinem
ent and verification of detector parameters during the mission and thus doe
s not depend on a priori knowledge of the system or ground calibration for
interpretation of images.