Abstract
Natural images are highly structured in their spatial configuration. It has been shown that the distribution of contrasts of natural images is almost always adequately described by a Weibull type distribution (Geuseboek & Smeulders, 2003) in which 2 free parameters are fitted. These 2 parameters appear to structure the natural image space in a highly meaningful way. Here we examined to which degree the brain is sensitive to these parameters by covarying the ERP responses of subjects viewing 1599 natural images with these 2 parameters. The fitted parameters explained 50% of the variance of the early ERP signal, contrasting sharply with more traditional contrast measurements like Fourier, Michelson and RMS, which explained only 5–20% of the ERP variance. In addition, we found a correlation of 0.84% and 0.93% between the two parameters of the Weibull type distribution fit and a model of the parvo- and magnocellular system. Both observations suggest that the parvo- and magnocellular system evolved, amongst others, to estimate these parameters.