search

markisus / pytagmapper

Python Mapping library for Fiducial Tags
MIT License
19 stars 2 forks source link

paper or some explain blogs #1

Open TE-HA opened 2 years ago

TE-HA commented 2 years ago

Thanks for your great work! I would like to know if there are any papers or some supplementary materials about this.

markisus commented 2 years ago

Thanks for checking out my project TE-HA. The implementation uses gaussian belief propagation. There is a nice tutorial https://gaussianbp.github.io/ In this project, each graph only has binary factors which simplifies matters greatly.

Each variable is an element of SE3 (3D pose) whose derivative is in se3. You can read about that here https://arxiv.org/abs/1812.01537 Each factor is a nonlinear function that takes a camera pose and a tag pose and computes the pixel error from the detected pixels.

Please let me know if you have any more specific questions.