This is a collection of educational notebooks on multi-view geometry and computer vision. Subjects covered in these notebooks include:
- Camera calibration
- Perspective projection
- 3D point triangulation
- Quaternions as 3D pose representation
- Perspective-n-point (PnP) algorithm
- Levenberg–Marquardt optimization
- Epipolar geometry
- Relative poses from stereo views (w. fundamental matrix)
- Bundle adjustment
- Structure from motion
The notebooks can be run in the browser by clicking the binder badge
# Builds the environment
docker build -t multiview_notebooks .
# Start a jupyter lab which can be opened in the browser
docker run -it --rm -p 8888:8888 multiview_notebooks jupyter-lab --ip=0.0.0.0 --port=8888
After starting the jupyter lab, the notebooks can be found in the home directory.
For the source of the Dockerfile, see this repository
