Human visual system can easily recognize a variety of surface states such as wetness of a bath floor, slipperiness of a road, or dustiness of an ornament. The perception of “Shitsukan” including the surface states is a critical missing piece in the sciences on biological and machine sensory processing, and recently gathers a great deal of interest. We have revealed the visual mechanisms underlying the wetness perception of a scene by integrating methodologies of human psychophysics, computer vision, computer graphics, and information theory. Furthermore, by applying the scientific finding to the image processing technology, we have developed a novel material transformation, the “wet filter”, that makes a dry object look wet. In the future, science-based material technologies like the present “wet filter” will give us a way to easily edit an object’s appearance, and let robots quickly recognize “Shitsukan” of an object just as human do.

Please click the thumbnail image to open the full-size PDF file.