Fast CGH computation using S-LUT on GPU
Data Storage Institute, A*STAR (Agency for Science, Technology and Research), DSI Building, 5 Engineering Drive 1, (Off Kent Ridge Crescent, NUS), Singapore 117608
Opt. Express, Vol. 17, No. 21. (12 October 2009), pp. 18543-18555.
@article{pan2009fast,
title={Fast cgh computation using s-lut on gpu},
author={Pan, Y. and Xu, X. and Solanki, S. and Liang, X. and Tanjung, R.B.A. and Tan, C. and Chong, T.C.},
journal={Optics Express},
volume={17},
number={21},
pages={18543–18555},
year={2009},
publisher={Optical Society of America}
}
In computation of full-parallax computer-generated hologram (CGH), balance between speed and memory usage is always the core of algorithm development. To solve the speed problem of coherent ray trace (CRT) algorithm and memory problem of look-up table (LUT) algorithm without sacrificing reconstructed object quality, we develop a novel algorithm with split look-up tables (S-LUT) and implement it on graphics processing unit (GPU). Our results show that S-LUT on GPU has the fastest speed among all the algorithms investigated in this paper, while it still maintaining low memory usage. We also demonstrate high quality objects reconstructed from CGHs computed with S-LUT on GPU. The GPU implementation of our new algorithm may enable real-time and interactive holographic 3D display in the future.
October 28, 2010 by hgpu
