Skip to content

Latest commit

 

History

History
14 lines (13 loc) · 1.6 KB

README.md

File metadata and controls

14 lines (13 loc) · 1.6 KB
Raw

TGV-GPU

TGV-GPU is a GPU implementation for the Total Generalized Variation (TGV) stereo algorithm [1]. Based on the original paper, for initialization, the author uses the Census transformation based locally adaptive support-weight aggregation [2] and WTA. For the anisotropic diffusion tensor computation [3] the implementation uses the consistent gradient operator [5]. Although this is a full GPU implementation edge-segment based adaptive regularization based on the LSD [4] is NOT implemented. For exhaustive search of regularized cost function the author adopts CUB's block reduce algorithm. The result is not so impressive, so any comments, bug fix and improvements are welcome.

Reference

  1. Kuschk, G., & Cremers, D. (2013). Fast and accurate large-scale stereo reconstruction using variational methods. In Proceedings of the IEEE International Conference on Computer Vision Workshops (pp. 700-707).
  2. Yoon, K. J., & Kweon, I. S. (2006). Adaptive support-weight approach for correspondence search. IEEE transactions on pattern analysis and machine intelligence, 28(4), 650-656.
  3. Werlberger, M., Trobin, W., Pock, T., Wedel, A., Cremers, D., & Bischof, H. (2009). Anisotropic Huber-L1 Optical Flow. In BMVC (Vol. 1, No. 2, p. 3).
  4. Von Gioi, R. G., Jakubowicz, J., Morel, J. M., & Randall, G. (2008). LSD: A fast line segment detector with a false detection control. IEEE transactions on pattern analysis and machine intelligence, 32(4), 722-732.
  5. S. Ando. Consistent gradient operators. IEEE Transactions on Pattern Analysis and Machine Intelligence, 22(3), 252-265. March 2000.