Tri-Focal Rig (Practical Camera Configurations for Image and Depth Acquisition)

In many VFX productions the existence of a disparity or depth map for the captured images is of the essence. This article describes two approaches for simultaneous acquisition of image and depth information. The first approach utilizes three cameras, one primary motion picture camera and two smaller auxiliary cameras. A robust and set-ready tri-focal camera rig was designed and implemented, which meets the requirements of motion picture professionals and enables precise and robust alignment of the cameras and perfect synchronization. — In the second approach a depth camera based on the principle of time-of-flight is integrated with a motion picture camera forming the so-called Motion Scene Camera. A rotating mirror shutter allows both sensors to utilize a single lens and capture the scene from the same viewing point. Both concepts have taken extensive tests, including a real production for the tri-focal approach and first field tests for the motion scene camera. Results are very encouraging and exalt the imagination of novel post production applications like hybrid 3D production and depth based masking and compositing.

Published

2013-10

Content type

Original Research

Keywords

Computational cinematography, depth map, disparity map, hybrid 3D, motion scene camera, multi-camera array, stereo, time-of-flight, tri-focal

DOI

10.5594/M001492

ISBN

978-1-61482-953-9