Estimating Light Vectors in Real Time

  • Authors:
  • Wai-Man Pang;Tien-Tsin Wong;Pheng-Ann Heng

  • Affiliations:
  • The Chinese University of Hong Kong;The Chinese University of Hong Kong;The Chinese University of Hong Kong

  • Venue:
  • IEEE Computer Graphics and Applications
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

The determination of the light vector is important for capturing the surface appearance under different illumination conditions. This is necessary for bi-directional reflectance distribution function measurement, texture acquisition, and image-based data acquisition. Most current systems adopt mechanical setup to control the light source position and direction. However, the setup is usually expensive and not portable. It also imposes limitation on the size of object being captured. This article proposes a capturing system that estimates the light vector by applying a vision-based technique. The user can freely move a handheld light source during capture. The light vector is estimated in real-time through a Web cam mounted on the light source. This application acquires image-based data for relighting. Also presented is a discussion on how the captured data is processed and compressed to a manageable form.