Perecption of Surface Textures

Posted 23 Apr 2007 at 18:38 UTC by steve

An MIT news release describes the underlying neural mechanism researcher believe is responsible for our ability to visually categorize surface textures. We can tell at a glance whether a surface is a rough or glossy. The researchers believe our brain has collections of neurons that respond to positive or negative skewing of a luminance histogram. The asymmetry of the histogram correlates to the surface quality. Glossy, darker surfaces tend to be skewed positively, while rougher surfaces are skewed negatively. There are immediately applications to machine vision, with autonomous vehicles high on the list. Being able to easily determine if the road is icy or dry, paved or dirt would be very useful. NTT, who collaborated on this research, offers a more detailed summary of the research which includes example images and histograms.

Specular Clues, posted 24 Apr 2007 at 17:10 UTC by Nelson » (Journeyer)

There are several visual clues that our eyes use to distinguish dull matt (labertian) reflective surfaces from shiny glossy specular reflective surfaces.

In addition to the dramatic intensity changes in the vicinity of highlights, we also have the advantage of eyeballing the surface from two different positions at the same time. Because shiny reflections tend to be very directional, these highlights tend to look very different from these two positions and are are quite noticeable.

To give you an idea of the effect, try using Red-Green stereo glasses to look at the room around you. You might notice that any red or green surfaces look distinctly shiny. This is because your eyes notice such a dramatic difference in brightness, which your brain tranlates as a shiny directional reflection.

See more of the latest robot news!