A comparative analysis is made of visual-perception models for estimating how the probability of distinguishing a target object is affected by background noise caused by the inhomogeneity of an outdoor scene and by the objects included in it. It is shown that a model that takes into account the regularities of the perception of visual stimuli and semantic objects describes the experimental data concerning the probability of distinguishing an object in the presence of background noise more simply and with greater reliability than a model based on signal-detection theory.
L. N. Aksyutov, "Analysis of models for estimating how background noise affects the probability of distinguishing objects visually," J. Opt. Technol. 70, 642-648 (2003)