Simple SLAM pipeline that creates a point cloud based map from a video input. This was implemented as part of a course project to understand and implement the data association and backend optimization of SLAM
The codebase is heavily borrowed from twitchslam with minor refactoring.
- OpenCV
- Visualization - Pangolin and python bindings pangolin
- Optimization - g2o and its python bindings g2opy
- skimage
python main -i /path/to/video -f focal-length