Visualization and evaluation

Visualization

There are .rviz config files for each algorithm (adapted from [https://github.com/engcang/vins-application](https://github.com/engcang/vins-application).

Additionally, you can visualize the trajectory using each algorithm's custom scene if you are running it live or plot the trajectory using evo

Evaluation

Evaluation of each algorithm's performance is done using evo. This python module can be used to plot trajectories with time and spatial alignment, calculate metrics such as APE and RPE and convert between formats like tum, euroc and bag files. As the module still has a lot of python2 dependencies it is advised to create a virtual environment to avoid python version conflicts:

# Get virtualenv package
apt-get install -y virtualenv

# Create and activate 
virtualenv -p python venv
source ./venv/bin/activate

# Install evo and necessary dependencies
cd /eval/evo
python -m pip install --editable . --upgrade --ignore-installed --no-binary evo
pip install --extra-index-url https://rospypi.github.io/simple/ rospy rosbag tf2_ros
pip install rospkg

There's really nothing more I can say that hasn't been well explained in the official repo wiki