Abstract
In this paper, we present a method for synthesizing virtual appearance of an object by projecting images onto the surface of the object using projectors. The object surface is assumed to have a known diffuse reflectance property; its shape is allowed to have an arbitrary shape. Using a system consisting of multiple projectors and a camera, the method first estimates their internal and external parameters as well as the object surface based on the projection of structured patterns, and then measures surface normals by the method of photometric stereo that uses the same projectors as point sources of illumination. By enabling highly accurate calibration of the projectors as well as reconstruction of the object shape and also by reducing the random errors in surface normals that significantly affect final appearance, it is made possible to synthesize high-quality appearance associated with an arbitrary virtual reflectance property.
Original language | English |
---|---|
Pages (from-to) | 410-422 |
Number of pages | 13 |
Journal | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
Volume | 5414 LNCS |
DOIs | |
Publication status | Published - 2009 |
Event | 3rd Pacific Rim Symposium on Image and Video Technology, PSIVT 2009 - Tokyo, Japan Duration: 2009 Jan 13 → 2009 Jan 16 |
Keywords
- Augmented reality
- Autocalibration
- Photometric stereo
- Projector-camera
ASJC Scopus subject areas
- Theoretical Computer Science
- Computer Science(all)