Real-Time Image Segmentation on a GPU
Georg-August University, Bernstein Center for Computational Neuroscience, Department for Computational Neuroscience , III Physikalisches Institut, Gottingen, Germany
Facing the MulticoreChallenge (2011) Volume: 6310, Publisher: Springer Berlin Heidelberg, Pages: 131-142-142
@article{abramov2011real,
title={Real-Time Image Segmentation on a GPU},
author={Abramov, A. and Kulvicius, T. and W{\”o}rg{\”o}tter, F. and Dellen, B.},
journal={Facing the Multicore-Challenge},
pages={131–142},
year={2011},
publisher={Springer}
}
Efficient segmentation of color images is important for many applications in computer vision. Non-parametric solutions are required in situations where little or no prior knowledge about the data is available. In this paper, we present a novel parallel image segmentation algorithm which segments images in real-time in a non-parametric way. The algorithm finds the equilibrium states of a Potts model in the superparamagnetic phase of the system. Our method maps perfectly onto the Graphics Processing Unit (GPU) architecture and has been implemented using the framework NVIDIA Compute Unified Device Architecture (CUDA). For images of 256×320 pixels we obtained a frame rate of 30 Hz that demonstrates the applicability of the algorithm to video-processing tasks in real-time.
March 13, 2011 by hgpu
