QPEP-DSO: Quadratic Pose Estimation Problems (QPEP) Enhanced Direct Sparse Odometry with Loop Closure

Code Arch

The codes are motivated by DSO (https://github.com/JakobEngel/dso) by Engel et al. and LDSO (https://github.com/tum-vision/LDSO) by Xiang Gao et al.

The compiler has to support C++14 standard.

Build

First, come to the thirdparty directory and then compile g2o and DBoW3 (DO NOT INSTALL!!!). Then, build the project. The final binaries locate at bin directory.

Trouble shooting:

1. Pangolin_FOUND is not TRUE: This means that the Pangolin cmake file did not change the enviroment variable. Use cmake -DPangolin_FOUND=TRUE to gap this error.

2. /usr/bin/ld: cannot find -lEigen3::Eigen: This means that the Pangolin cmake file has some problems, try this command and then rebuild: sudo ln -s /usr/local/lib/libpango_core.so /usr/local/lib/libEigen3::Eigen.so

Related Publications

• LDSO: Direct Sparse Odometry with Loop Closure, X. Gao, R. Wang, N. Demmel, D. Cremers, In International Conference on Intelligent Robots and Systems (IROS), 2018.
• Direct Sparse Odometry, J. Engel, V. Koltun, D. Cremers, In IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), 2018
• A Photometrically Calibrated Benchmark For Monocular Visual Odometry, J. Engel, V. Usenko, D. Cremers, In arXiv:1607.02555, 2016
• **Quadratic Pose Estimation Problems: Globally Optimal Solutions, Solvability/Observability Analysis and Uncertainty Description **, J. Wu, Y. Zheng, Z. Gao, Y. Jiang, X. Hu, Y. Zhu, J. Jiao, M. Liu, 2020, https://github.com/zarathustr/QPEP

Run

• TUM-Mono Dataset: https://vision.in.tum.de/mono-dataset

To run QPEP-DSOP on TUM-Mono dataset sequence 34, execute:

./bin/run_dso_tum_mono \
    preset=0 \
    files=XXXXX/TUMmono/sequences/sequence_34/images.zip \
    vignette=XXXXX/TUMmono/sequences/sequence_34/vignette.png \
    calib=XXXXX/TUMmono/sequences/sequence_34/camera.txt \
    gamma=XXXXX/TUMmono/sequences/sequence_34/pcalib.txt \
    method=QPEP-RANSAC