OpenGV
A library for solving calibrated central and non-central geometric vision problems
relative_pose Directory Reference

Files

file  CentralRelativePoseSacProblem.hpp [code]
 Functions for fitting a relative-pose model to a set of bearing-vector correspondences, using different algorithms (central). Used with a random sample paradigm for rejecting outlier correspondences.
 
file  EigensolverSacProblem.hpp [code]
 Functions for fitting a relative-pose model to a set of bearing-vector correspondences, using the eigensolver algorithm. Used with a random sample paradigm for rejecting outlier correspondences.
 
file  MultiCentralRelativePoseSacProblem.hpp [code]
 Functions for fitting a multi relative-pose model to sets of bearing-vector correspondences between multiple viewpoints. Exploits mutual agreement. Used within a multi random-sample paradigm for rejecting outlier correspondences. Experimental!
 
file  MultiNoncentralRelativePoseSacProblem.hpp [code]
 Functions for fitting a relative-pose model to a set of bearing-vector correspondences from a multi-camera system (non-central). Uses RelativeMultiAdapterBase for managing correspondence-lists between individual pairs of views, enabling homogeneous sampling over the views. Used in a random-sample paradigm for rejecting outlier correspondences.
 
file  NoncentralRelativePoseSacProblem.hpp [code]
 Functions for fitting a relative-pose model to a set of bearing-vector correspondences from a multi-camera system (non-central). Used with a random-sample paradigm for rejecting outlier-correspondences.
 
file  RotationOnlySacProblem.hpp [code]
 Functions for fitting a rotation-only model to a set of bearing-vector correspondences. The viewpoints are assumed to be separated by rotation only. Used within a random-sample paradigm for rejecting outlier correspondences.
 
file  TranslationOnlySacProblem.hpp [code]
 Functions for fitting a relative-pose model to a set of bearing-vector correspondences. The viewpoints are assumed to be separated by translation only. Used within a random-sample paradigm for rejecting outlier correspondences.