Modeling spectrally varying resolution in broadband imaging systems

Abstract

The remote sensing system engineering process often makes use of modeling and simulation tools to flow down specifications to subsystems and components, and/or to predict performance given a particular set of defined capabilities. A persistent question in the development and use of such tools is that of appropriate level of fidelity. In this paper we look at one problem area encountered in the simulation of panchromatic and other broadband imaging systems, that of accounting for spectrally varying resolution over the band. An established method for treating this variation is that of the polychromatic optical transfer function (OTF), but this technique imposes a measure of complexity on the simulation tool software architecture, as well as on users who must subsequently interact with it. We present a methodology for assessing the required level of fidelity for this problem and show that under some conditions it appears possible to forgo the polychromatic OTF formalism, or else to treat it with less than full rigor, with minimal loss in accuracy.

Simple techniques for modeling wavefront error in imaging systems

“…In particular, for both the phase map and OQF approximations, we calculate the relative edge response (RER), which is a key term in the GIQE. The GIQE, NIIRS scale, and the RER are all explained in further detail in a companion to this paper [7]. We then use the GIQE to determine the difference in the NIIRS scores that would be predicted for the phase map and OQF approximations to the effect of wavefront error.…”

“…A more complete description of ensquared energy and its utility for describing the performance of an optical system is given in [7]. In this study we consider the case of a Q=1 system and use a Monte Carlo simulation to capture the diversity of ensquared energies calculated via the phase map method and to compare the phase map and OQF methods.…”