Covariance estimation for minimal geometry solvers via scaled unscented transformation

Evren Imre, Adrian Hilton

Abstract

Highlights•Scaled unscented transformation (SUT) for minimal geometry solvers (MGS).•SUT vs. first-order-propagation (FOP) in uncertainty estimation of MGS.•Since MGS are highly nonlinear, SUT is a better uncertainty estimator than FOP.
BibTeX
@article{imre2015covariance,
  author = {Evren Imre and Adrian Hilton},
  title = {{Covariance estimation for minimal geometry solvers via scaled unscented transformation}},
  journal = {Comput. Vis. Image Underst.},
  year = {2015},
  url = {https://openreview.net/forum?id=TwIEuBJv18},
}

Loading related content...

Figure