A hybrid approach to non-uniformity correction of large format emitter arrays

Abstract

Non-uniformity correction (NUC) of emitter arrays is an important part of the calibration of an infrared scene projector (IRSP), necessary to provide precise and artifact-free simulations. Producing an accurate and cost effective NUC of an IRSP is a challenge due to the complexity of the NUC process and the expense of high performance, large format infrared cameras. Previous NUC methods have typically fallen into either the sparse grid method or the flood method. The sparse grid method gives independent measurements of each emitter pixel, however, it is time consuming and becomes impractical for accurate measurements at low radiance levels, especially with lower performance but less expensive cameras such as microbolometers. Flood measurements are fast and can be applied at lower radiance, but do not allow precise measurement of the output of an individual pixel. Santa Barbara Infrared (SBIR) has developed a hybrid approach that makes use of both methods. Sparse grid methods are used at higher radiance levels to perform an initial NUC of the array. Then, a combination of flood and sparse grid data is used to extend the NUC to lower radiance levels and improve the high radiance NUC through iteration. Details of the approach and results from its application to an emitter array are presented.