Noise-Limited Imaging and Target Acquisition

doi:10.1117/3.2265069.ch11

Book: A System Engineering Approach to Imaging

Author(s): Norman S. Kopeika

Published: 1998

https://doi.org/10.1117/3.2265069.ch11

Author Affiliations +

Abstract

Basic noise concepts were developed in Chapter 5 and applied to individual “point” detectors in Chapter 6. Here, these concepts are extended to the image as a whole, where object or target area must be considered. Noise manifests itself when irradiance received from the object plane scene is weak, for example, as random “snow” when weak TV signals are received. This stems from random time and space fluctuations in image irradiance giving rise to random nonuniformities in image brightness, and also from electronic noise. In noise-limited imaging, the object or target contrast is imaged against noise, and the latter rather than required threshold contrast limits the ability of the observer to resolve the object or target. The poorer the signal-to-noise ratio (SNR), the poorer the image quality and resolution. Quantitatively, system modulation transfer function (MTF) is incorporated into the expression for signal power, thus giving rise to SNR as a function of spatial frequency. Intersection of system SNR curve with required SNR threshold defines fr max for noise-limited imaging, as shown in Fig. 11.1.