Bilateral Filtering with CUDA
University of Aarhus
University of Aarhus, 2012
@article{staal2012bilateral,
title={Bilateral Filtering with CUDA},
author={Staal, Lasse Klojgaard},
year={2012}
}
This paper implements the Bilateral filter, using CUDA enhanced parallel computations. The Bilateral filter allows smoothing images, while preserving edges, in contrast to e.g. the Gaussian filter, which smoothes across edges. While delivering visually stunning results, Bilateral filtering is a costly operation. Using NVidia’s CUDA technology the filter can be parallelized to run on the GPU, which allows for fast execution, even for high definition images.
January 27, 2012 by hgpu
