Silhouette Refining for the Volume Intersection Method with Random Pattern Backgrounds

Masahiro TOYOURA  Masaaki IIYAMA  Koh KAKUSHO  Michihiko MINOH  

D - Abstracts of IEICE TRANSACTIONS on Information and Systems (Japanese Edition)   Vol.J89-D   No.11   pp.2413-2424
Publication Date: 2006/11/01
Online ISSN: 1881-0225
Print ISSN: 1880-4535
Type of Manuscript: PAPER
volume intersection method,  silhouette refining,  color estimation,  random pattern background,  

Full Text(in Japanese): PDF(781.6KB)
>>Buy this Article

The volume intersection method reconstructs shapes of objects from their silhouettes obtained by multiple cameras. There occurs a problem when some parts of the silhouettes are missing; some parts of reconstructed shapes are missing. We propose a method for refining silhouettes to reconstruct more accurate shapes. Random pattern background keeps missing parts of silhouettes below a certain rate. Once a rough shape is reconstructed, the object's color is estimated with images of multiple cameras. Missing parts can be refined by comparing the object's color with its corresponding background's color. In our experiment, proper silhouette refining was confirmed by comparison between the reconstructed shape and the shape from manually extracted silhouettes.