PROCEEDINGS ARTICLE | May 3, 2017
Proc. SPIE. 10178, Infrared Imaging Systems: Design, Analysis, Modeling, and Testing XXVIII
KEYWORDS: Point spread functions, Visual process modeling, Imaging systems, Sensors, Video, Interference (communication), 3D modeling, Night vision, Image analysis, Quantization, Integration, Modulation transfer functions, Night vision systems, Performance modeling, Systems modeling
Characterizing an imaging system through the use of linear transfer functions allows prediction of the output for an arbitrary input. Through careful measurement of the systems transfer function, imaging effects can then be applied to desired imagery in order to conduct subjective comparison, image based analysis, or evaluate algorithm performance. The Night Vision Integrated Performance Model (NV-IPM) currently utilizes a two-dimensional linear model of the systems transfer function to emulate the systems response and additive signal independent noise. In this correspondence, we describe how a two-dimensional MTF can be obtained through correct interpolation of one-dimensional measurements. We also present a model for the signal dependent noise (additive and multiplicative) and the details of its calculation from measurement. Through modeling of the experimental setup, we demonstrate how the emulated sensor replicates the observed objective performance in resolution, sampling, and noise. In support of the reproducible research effort, many of the Matlab functions associated with this work can be found on the Mathworks file exchange [1].
