Computational simulation of human perception of spatially-dependent patterns modulated by degree and angle of linear polarization

Abstract

Recent studies on polarization perception have shown that humans are sensitive to patterned stimuli modulated by either angle of linear polarization (AoP) or degree of polarization (DoP). Here, we present a model of human polarization sensitivity that incorporates both AoP and DoP as spatially dependent input variables. Applying the model to both sinusoidal- and square-wave-modulated DoP and AoP inputs, we demonstrate the theoretical similarities and differences generated by such inputs. Our model indicates the following: (i) edge boundaries between two adjacent areas of different linear polarization are preserved for both AoP- and DoP-modulated stimuli; and (ii) compared with DoP stimuli, AoP stimuli generate greater luminance changes at the photoreceptor level, suggesting that AoP-modulated patterns are potentially more salient than DoP patterns. The computational model is supported experimentally with an optical test of the model comprising a radial diattenuating polarizing filter and modified liquid crystal displays generating DoP- and AoP-modulated outputs. Psychophysical measures of human sensitivity confirm the increased salience of AoP- relative to DoP-modulated stimuli. These findings have practical application to the selection of DoP- and AoP-modulated stimuli for the investigation of macular function and macular pigment density in healthy and diseased eyes.