Method of determining probability of target detection in a visually cluttered scene

摘要

A method to determine the probability of detection, P(t), of targets within infrared-imaged, pixelated scenes includes dividing the scenes into tar blocks and background blocks. A plurality of &Dgr;T metrics are applied to the blocked scenes to derive various &Dgr;T values for each scene. Factor analysis is then used on the various . DELTA.T values to derive a relative &Dgr;T for each scene. The scenes are divided again, into cells of pixels, in accordance with a plurality of clutter metrics. These clutter metrics are used to derive various clutter values for each scene. Factor analysis is used on the various clutter values to derive relative clutter values for each scene. The relative &Dgr;T values and the relative clutter values are used in determining the probabilities of detection of the targets in the scenes. Based on the probabilities of detection, the optimum scene or set of scenes is selected.