Experimental method for observation prediction based on the decision matrix, through day/ night equipments in NIR and LWIR spectral ranges

摘要

The paper presents an evaluation methodology and the results of some experiments that have been made in laboratory in order to determine the target's detection's probability depending on the target's contrast and the observers' age. The main goal was to assure the model for an optimal feature's configuration for a device used to enable the view during day or night, so that we can estimate, within improper view conditions, its visibility boundaries during day and night. The base of method's principle is the Bayes' theorem, and the authors have used in their experiments the technique of estimation by probability of real positive and real negative that is also used in medical evaluation of images. The authors have used an instrument layout in the laboratory that included an uncooled 8- 12 μm thermal camera, a CCD and a ICU camera, an USAF pattern and a set of chemical compositions that produce aerosols with different concentrations. It has been proved that the detection probability decreases proportionally by age, but being differentiated by the contrast between the target and the background; it has been presented the diagram of the probability variation and the analytical relationships that approximate it, in terms of contrast and aerosols' concentration features.