GPU-Based Foreground-Background Segmentation Using an Extended Colinearity Criterion
Swiss Federal Institute of Technology (ETH), Computer Vision Lab, Zurich, Switzerland
Proceedings of Vision, Modeling, and Visualization (VMV) 2005
@inproceedings{eth_biwi_00364,
author={Andreas Griesser and Stefaan De Roeck and Alexander Neubeck and Luc Van Gool},
title={GPU-Based Foreground-Background Segmentation using an Extended Colinearity Criterion},
booktitle={Proceedings of Vision, Modeling, and Visualization (VMV) 2005},
year={2005},
month={November},
pages={319-326},
editor={G. Greiner and J. Hornegger and H. Niemann and M. Stamminger},
publisher={IOS Press},
keywords={foreground-background segmentation, real-time, graphics hardware, gpu}
}
We present a GPU-based foreground-background segmentation that processes image sequences in less than 4ms per frame. Change detection wrt. the background is based on a color similarity test in a small pixel neighbourhood, and is integrated into a Bayesian estimation framework. An iterative MRF-based model is applied, exploiting parallelism on modern graphics hardware. Resulting segmentation exhibits compactness and smoothness in foreground areas as well as for inter-frame temporal contiguity. Further refinements extend the colinearity criterion with compensation for dark foreground and background areas and thus improving overall performance.
February 26, 2011 by hgpu
