The meditation event consists of a single person sitting on a small office table inside the Virtualized Reality project's 3D Dome. The data set consists of one frame for each of 51 cameras. Each frame is a 245x320-pixel image, with 8-bit greyscale pixels. The top few rows contain the VITC time code, which was used to help synchronize the video across the 51 cameras.
In addition to the images, we have also provided the camera calibration parameters that we use to represent the projection of the 3D world into the images. These models are generated by Reg Willson's freeware implementation of Roger Tsai's camera calibration algorithm.
All 51 images and (corrected) Tsai camera models. (3 MBytes)
NOTE: If you downloaded this data set before 19 MAY 1998, then the camera models are not correct. You can get just the 51 corrected camera models. (5368 bytes) (The latest data set includes a file "Version", which lists the data set as version 1.1. The initial release - version 1.0 - did not include this file.)