2024-12-11 23:50:27 -07:00
|
|
|
# VIO - PhotonVision
|
|
|
|
|
This is a project that aims to improve FRC vision accuracy using Visual Odometry (VO) and similar methods
|
|
|
|
|
|
|
|
|
|
Things to accomplish:
|
|
|
|
|
- [x] Get custom pipeline running
|
2024-12-15 15:54:25 -07:00
|
|
|
- [needs debugging] Basic VO (look at https://github.com/thehummingbird/robotics_demos/blob/main/monocular_vo)
|
2024-12-11 23:50:27 -07:00
|
|
|
- [ ] Inertial odometry using pidgen and sensor fusion (look at https://www.thinkautonomous.ai/blog/visual-inertial-odometry/)
|
2024-12-15 15:54:25 -07:00
|
|
|
- [?] Sensor fusion for physical odometry from RoboRiO
|
2024-12-11 23:50:27 -07:00
|
|
|
- [ ] Mesh based localisation? https://youtube.com/watch?v=vMsV04emXHU https://dlnext.acm.org/doi/10.1007/978-3-031-20047-2_34
|
|
|
|
|
- [ ] Optimise and refine.
|
