1. Overview

This code is an implementation of the following paper. The core solvers are written by C++. An Matlab Mex interface is provided for easy-to-use.

@article{zhao2020minimal,
  title={Minimal Case Relative Pose Computation Using Ray-Point-Ray Features},
  author={Zhao, Ji and Kneip, Laurent and He, Yijia and Ma, Jiayi},
  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
  year={2020},
  volume={42},
  number={5},
  pages={1176--1190}
}

In this paper, we proposed a few solvers for relative pose estimation from ray-point-ray (RPR) features.

We also proposed a few new five-point methods for relative pose estimation, which is a classical problem in geometric vision.

Note: We have tried 4 different rotation representations, and found that the Cayley representation has the best performance. Hence, only the solvers with the Cayley representation are provided.

Authors: Ji Zhao

2. Quick Start

Run compile.m in subfolders of "solvers_5pt_rpr" to compile the mex files. Make sure to set the proper path of Eigen library in compile.m. This step can be skipped once the mex files have been compiled. Compiled files using (Ubuntu 16.04 + Matlab R2019a) and (Windows 7 + Matlab 2016b + Visual Studio 2017) are provided.

Run test_solver*.m in folder "test"