my recent articles |
| Brain takes statistical approach to gauging glossiness |
|
20 April 2007, for Nature Network Boston How do we tell whether a surface is glossy or matte, wet or dry? According to a paper published online in Nature this week, our brains likely use a simple rule, based on the balance of light and dark spots seen on the surface, to reliably detect differences in surface quality. The researchers, including Lavanya Sharan and Edward Adelson of MIT, made images of stucco material painted with matte or glossy finishes, and digitally modified scans of Michelangelo sculptures to make them appear glossy or matte. They adjusted the images so that the total amount of light in each was the same.
Researchers showed subjects glossy and matte images to figure out how their brains can tell the difference. Credit: Digital Michelangelo Project They asked research subjects to rate the glossiness in these images. And they plotted the number of light and dark pixels in the images to show their distribution. For matte surfaces, the distribution of pixels was skewed towards the dark end of the spectrum, and for glossy surfaces, the distribution slanted towards the bright end. The shape of the distribution correlated strongly with the subjects’ glossiness ratings, the study showed, suggesting that the brain is sensitive to fine differences in the images’ lightness and darkness parameters. The same results were seen when subjects were looking at bits of crumpled paper or cotton fabric, suggesting that a similar mechanism in the brain underlies how we perceive a wide variety of textures. |
