Stereovision On GPU
University of Kentucky
EDGE Workshop (2006)
@conference{yang2006stereovision,
title={Stereovision on GPU},
author={Yang, R. and Wang, L. and Welch, G. and Pollefeys, M.},
booktitle={Workshop on Edge Computing Using New Commodity Architectures},
year={2006},
organization={Citeseer}
}
Depth from stereo has traditionally been, and continues to be one of the most actively researched topics in computer vision. Recent development in this area has significantly advanced the state of the art in terms of quality. However, in terms of speed, these best stereo algorithms typically take from several seconds to several minutes to generate a single disparity map, limiting their applications to off-line processing. There are many interesting applications, such as robot navigation and augmented reality, in which high-quality depth information at video rate is crucial.
March 21, 2011 by hgpu
