Light emitted from a light source, reaches to our eyes, after reflecting on a surface, refracting as it enters/leaves on a glass, and scattering inside a transparent object in a scene. By just looking an object, we, human, are able to recognize not only the surface physical condition (i.e. roughness and shininess) but also various kinds of object properties to decide the impression of the object. Thus, light must be a medium which transports various kinds of significant information of the scene. Our challenge is to develop an interface, which is named "Optical Media Interface", to share with information of the scene between human and computer according to the light transportation. To provide computers the ability of scene understanding from the "Optical Media Interface" is the aim of our research.

Computer Vision

We analyze the behavior of the light reflection on a surface, and scattering inside an object. This analysis should be a basic technology of material estimation and 3d shape reconstruction.

Computer Graphics and Feeling of Material

To reveal a mechanism of human visual perception, we develop human visual interface to represent reliable material appearance and feelings.

Computational Photography

We develop a method to generate images according to the computation of the light distribution captured by a specialized camera.To compute the distribution of the light captured by a specialized camera, we develop a method to generate an image which cannot be obtained from a conventional camera. Using this computation, we are able to generate an image taken by simulated virtual camera with physically impossible settings.This enables us to provide an image which we have never seen before.

Development of Sensing System

A Light transportation in 3D space contains various kinds of significant information, but a camera can only captures a 2D image of the scene which is just a subset of the light transportation. Therefore, constructing optical measurement devices with specialized mirrors and lenses, we challenge to capture the light transportation.

To page top