DM-VIO安装与运行各类数据集

1、论文地址：https://vision.in.tum.de/_media/research/vslam/dm-vio/dm-vio.pdfhttps://vision.in.tum.de/_media/research/vslam/dm-vio/dm-vio.pdf

2、源码地址：GitHub - lukasvst/dm-vio: Source code for the paper DM-VIO: Delayed Marginalization Visual-Inertial Odometryhttps://github.com/lukasvst/dm-vio

3、环境配置安装编译

        3.1 下载源码：

git clone https://github.com/lukasvst/dm-vio.git

        3.2 安装依赖： 

sudo apt-get install cmake libsuitesparse-dev libeigen3-dev libboost-all-dev libyaml-cpp-dev

        3.3 安装GTSAM

sudo apt install libtbb-dev
git clone https://github.com/borglab/gtsam.git
cd gtsam
git checkout 4.2a6          # not strictly necessary but this is the version tested with.
mkdir build && cd build
cmake -DGTSAM_POSE3_EXPMAP=ON -DGTSAM_ROT3_EXPMAP=ON -DGTSAM_USE_SYSTEM_EIGEN=ON -DGTSAM_BUILD_WITH_MARCH_NATIVE=OFF ..
make -j
sudo make install

        3.4 安装OpenCV

sudo apt-get install libopencv-dev

         3.5 Pangolin v0.6安装

sudo apt install libgl1-mesa-dev libglew-dev pkg-config libegl1-mesa-dev libwayland-dev libxkbcommon-dev wayland-protocols
git clone https://github.com/stevenlovegrove/Pangolin.git
cd Pangolin
git checkout v0.6
mkdir build
cd build
cmake ..
cmake --build .
sudo make install

 4 编译源码

cd dm-vio
mkdir build
cd build
cmake ..
make -j

 5 使用作者脚本运行数据集

        5.1 配置脚本依赖

pip3 install tqdm ruamel.yaml pyquaternion matplotlib tabulate

         5.2 下载脚本源码以及创建配置文件

git clone https://github.com/lukasvst/dm-vio-python-tools.git
cd dm-vio-python-tools
python3 create_config.py --name=workpc --dmvio_folder=/path/to/dmvio --results_folder=/results/will/be/saved/here

注： /path/to/dmvio     改成编译好的DM-VIO的路径。     

/results/will/be/saved/here   在编译好的DM-VIO路径下创建一个result文件夹用来保存运行好的数据集的相机轨迹文件。此处填写创建好的result的路径。

        5.3 下载你所需的数据集

python3 download_euroc.py --folder=/dataset/shall/be/saved/here/euroc
python3 download_tumvi.py --folder=/dataset/shall/be/saved/here/tumvi
python3 download_4seasons.py --folder=/dataset/shall/be/saved/here/4seasons

# Useful For testing: use --only_seq to only download the first sequence
python3 download_euroc.py --folder=/dataset/shall/be/saved/here/euroc --only_seq=0 

若不想下载所有数据集，可以加 --only_seq 然后加上你想要下载的数据集序列，详细序列可以查看 configs.yaml文件里的数据集序列排序。

6 数据集上运行DM-VIO

python3 run_dmvio.py --output=console --dataset=euroc --dmvio_settings=euroc.yaml --withgui --iter=1 --only_seq=10
python3 run_dmvio.py --output=console --dataset=tumvi --dmvio_settings=tumvi.yaml --withgui --iter=1 --only_seq=25
python3 run_dmvio.py --output=console --dataset=4seasons --dmvio_settings=4seasons.yaml --withgui --iter=1 --only_seq=0

 在这上一个跑了EuRoc数据集的效果图