Our paper is recived by IROS2025.

Figure 1: System framework of GeoFlow-SLAM.

Figure 2: Illustration of dual-stream optical flow estimation.

Figure 3: Structure of the factor graph used in optimization.

We have tested the library on Ubuntu 22.04 and 20.04, but it should be easy to compile on other platforms. A powerful computer (e.g., i7) will ensure real-time performance and provide more stable and accurate results.

We use the new thread and chrono functionalities of C++17.

We use Pangolin for visualization and user interface. Download and installation instructions can be found at: https://github.com/stevenlovegrove/Pangolin.

We use OpenCV to manipulate images and features. Download and installation instructions can be found at: http://opencv.org. Requires at least 3.0. Tested with OpenCV 3.2.0 and 4.4.0.

Required by g2o (see below). Download and installation instructions can be found at: http://eigen.tuxfamily.org. Requires at least 3.1.0.

We use modified versions of the DBoW2 library to perform place recognition and the g2o library to perform non-linear optimizations. Both modified libraries (which are BSD) are included in the Thirdparty folder.

The modified GMS and small_gicp are used.

We provide a script build.sh to build the Thirdparty libraries and geoflow-slam. Please make sure you have installed all required dependencies (see section 2). Execute:

cd geoflow-slam
chmod +x build.sh
./build.sh

This will create libORB_SLAM3.so in the lib folder and the executables in the Examples folder.

cd geoflow-slam/script/tools
bash parse_d435i.sh

cd geoflow-slam/script/run_orbslam
python run_rgbd_vi_g1.py

The ROS2 node is also provided in Examples/ROS2/RGB-D-Inertial. The robust Mono-Inertial version tested in gopro camera is also provided in Examples/Monocular-Inertial.

The datasets are released to unitree_legged_robotic_datasets. Thank you for citing our paper.

@misc{xiao2025geoflowslamrobusttightlycoupledrgbdinertial,
      title={GeoFlow-SLAM: A Robust Tightly-Coupled RGBD-Inertial and Legged Odometry Fusion SLAM for Dynamic Legged Robotics}, 
      author={Tingyang Xiao and Xiaolin Zhou and Liu Liu and Wei Sui and Wei Feng and Jiaxiong Qiu and Xinjie Wang and Zhizhong Su},
      year={2025},
      eprint={2503.14247},
      archivePrefix={arXiv},
      primaryClass={cs.RO},
      url={https://arxiv.org/abs/2503.14247}, 
}