Image structural difference clutter metric based on human visual properties

Optoelectronic image clutter is seriously affecting target acquisition performance, in this paper, an image structural difference clutter metric based on human visual properties was proposed for quantifying background clutter. On the basis of high adaptability of human visual perception in extracting image structural information, a structural similarity measuring map was calculated between the target and clutter images according to the widely accepted definition of the structural-similarity (SSIM) index in the field of image quality assessment(IQA). And then, based on human vision saliency, two kinds of suitable saliency information weighting measures were developed to pool the structural similarity measuring map into a metric. The metric was tested in an outfield experiment provided by TNO Human Factors Research Institute of Netherlands. Comparative experiments demonstrate that this metric makes an improvement in RMS error and correlation than previously proposed image clutter metrics in predicting target acquisition performances including detection probabilities and false alarm probabilities.