Photo-Realistic 3D Model Extraction from Camera Array Capture

In which the authors describe a process of capturing photo-realistic 3D computer models using purely passive methods based on subject capture with arrays of cameras, and image analysis to permit the instant capture of both form and texture of the subject or talent. — The key performance characteristics of the array are discussed, particularly the challenges of triggering, and the limits imposed on the content created by the native resolution of the cameras used in the array. — Details of the rig design and camera layout and configuration for efficient and effective subject capture are presented. — The process by which multiple still photographs are processed to produce a point cloud which in turn becomes the model mesh are presented, together with examples of the current state of the art of this approach. — Tradeoffs such as the decision to eschew the use of active techniques such as laser scanning, structured light projection, or time-of-flight techniques are discussed, together with their benefits.

Published

2011-06

Content type

Original Research

Keywords

3D scanning, Full Body, Camera array, Photometrics, Passive scanning, Sculptural Photography, Free Viewpoint Media

DOI

10.5594/M001423

ISBN

978-1-61482-951-5