posest - Robust 6DoF Pose Estimation from 3D-2D Correspondences
This is posest, a library for 3D pose estimation from point correspondences that is distributed under the GPLv2. Pose estimation refers to the computation of position and orientation estimates that fully define the posture of a rigid object in space (6 DoF in total). The estimation is based on a set of known 3D points and their corresponding 2D projections on an image.
The library estimates the relative motion between the 3D points and the camera system. Single or binocular camera systems are supported. Image points typically originate from local features, however posest is oblivious to their origin. Mismatched 3D-2D point pairs (i.e. outliers) are handled with RANSAC. More details can be found at posest's web site.
posest requires the levmar library to be built. By default, levmar requires LAPACK or an equivalent library. However, it can also be configured to use a LAPACK-free linear systems solver.
After installing the required dependencies, create a build directory in the root of the cloned repository and run cmake.
posest has a matlab mex interface in the matlab subdirectory. See the included README.txt for more information.
If you use this code in your published work, please cite the following paper:
@incollection{Lourakis13model,
author={Lourakis, Manolis and Zabulis, Xenophon}
title="{Model-Based Pose Estimation for Rigid Objects}",
booktitle={Computer Vision Systems},
isbn={978-3-642-39401-0},
volume={7963},
series={Lecture Notes in Computer Science},
editor={Chen, Mei and Leibe, Bastian and Neumann, Bernd},
publisher={Springer Berlin Heidelberg},
doi={10.1007/978-3-642-39402-7_9},
pages={83--92},
year="2013",
url="http://users.ics.forth.gr/~lourakis/posest/ICVS13.pdf"
}