A Contrast/Filling-In Model of 3-D Lightness Perception
Ross, William D.
MetadataShow full item record
Wallach's ratio hypothesis states that local luminance ratios clr!termine lightness perception under variable illumination. While local luminance ratios successfully discount gradual variations in illumination (illumination constancy or Type I constancy), they fail to explain lightness constancy in general. Some examples of failures of the ratio hypothesis include effects suggesting the coplanar ratio hypothesis (Gilchrist 1977), "assimilation" effects, and configural effects such as the Benary cross, and White's illusion. The present article extends the Boundary Contour System/Feature Contour System (BCS/FCS) approach to provide an explanation of these effects in terms of a neural model of 3-D lightness perception. Lightness constancy of objects in front of different backgrounds (background constancy or Type II constancy) is used to provide functional constraints to the theory and suggest a contrast negation hypothesis which states that ratio measures between coplanar regions are given more weight in the determination of lightness. Simulations of the model applied to several stimuli including Benary cross and White's illusion show that contrast negation mechanisms modulate illumination constancy mechanisms to extend the explanatory power of the model. The model is also used to devise new stimuli that test theoretical predictions.