With the increased technological advances of thermal imagers, existing theoretical models are not capable of predicting the advantages of advanced thermal imagers. The advanced thermal imager contains a staring focal plane array, which may be sensitive to multiple spectra, and light enough to be mounted on a rifle. Research was conducted to address portions of these three areas. Psychophysical experiments were conducted to model the impact of insufficient sampling on human observer performance for the identification of military vehicles. An information comparison metric was proposed for the comparison of different spectral images. The impact of dead pixels and thermal imager noise on the comparison metric were measured. A human performance model was then developed to predict a humans ability to identify small handheld objects. The impacts of these models on current and future research were discussed.