High-Speed Dense Stereo Via Directional Center-Biased Windows on Graphics Hardware

Jiangbo Lu, S. Rogmans, G. Lafruit, F. Catthoor
Leuven Univ., Leuven
3DTV Conference, 2007


   title={High-speed dense stereo via directional center-biased windows on graphics hardware},

   author={Lu, J. and Rogmans, S. and Lafruit, G. and Catthoor, F.},

   booktitle={3DTV Conference, 2007},





Source Source   



We present a high-speed dense stereo algorithm that achieves both good quality results and very high disparity estimation throughput on the graphics processing unit (GPU). The key idea is to make use of directional center-biased support windows to strike a good quality balance between homogeneous areas and depth discontinuities, and it is instantiated by a truncated separable Laplacian kernel approximation. To boost its real-time speed and tackle the computation precision problem on GPUs, a number of critical optimization schemes have been proposed. Evaluation using the benchmark Middlebury stereo database shows that our approach is better than previous GPU-based local stereo methods and it even outperforms some methods using global optimization techniques. Our optimized implementation completely running on an Nvidia GeForce 7900 graphics card achieves over 668 million disparity estimations per second (Mde/s) including all the overhead, which is about 2.3 to 13.4 times faster than the GPU-based solutions ever reported.
No votes yet.
Please wait...

* * *

* * *

HGPU group © 2010-2021 hgpu.org

All rights belong to the respective authors

Contact us: