Analysis and implications of resistive emitter array non-uniformity correction (NUC) between sensors with different spectral bands

Resistive array technology is finding increasing application in representing synthetic infrared targets and backgrounds. This paper examines the implications for resistive array calibration and radiance nonuniformity when using a NUC sensor different from the intended unit under test (UUT). Pixel-to- pixel radiance nonuniformity is the prominent noise source from resistive arrays and must be minimized for high-fidelity testing of infrared imaging sensors. We begin by analyzing three primary concerns that arise from using a NUC sensor other than the UUT: differences in responsivity and transmission of the two sensors, and the spectral emissivity of the resistive array. Because the emitters are not blackbodies, one must introduce a model of the emittance that is more complicated than Planck's function alone and involves knowledge of the emitter spectral emissivity. Sensor responsivity and transmission may further emphasize one portion of the spectral band over another. Next, we discuss how these parameters can lead to nonuniformity for the UUT even when resistive array output is perfectly calibrated for the NUC sensor. We also examine how these three parameters impact band contrast and absolute radiometry of a projected scene. We propose a method that has been used in the past to provide radiometrically accurate signals (in equivalent blackbody temperature units) to the UUT. The paper concludes by analyzing the practical side of the proposed method and establishes error bounds on the procedure as being implemented on the U.S. Army's Dynamic Infrared Scene Projector (DIRSP).