Keywords: Superpoint Feature detection, SuperGlue Feature matching, Five-Point Motion Estimation

port from https://github.com/yueying/LearningVO change the traditional feature dectection and matching to learning based

Requirements

• Python 3
• Numpy
• OpenCV

Dataset

KITTI odometry data set (grayscale, 22 GB)

Usage

Modify the path in test.py to your image sequences and ground truth trajectories, then run

python test.py

References

1. 一个简单的视觉里程计实现 | 冯兵的博客
   
2. Monocular Visual Odometry using OpenCV and its related project report Monocular Visual Odometry | Avi Singh
3. uoip/monoVO-python
4. magicleap/SuperGluePretrainedNetwork

Search "cv2.findEssentialMat", "cv2.recoverPose" etc. in github, you'll find more python projects on slam / visual odometry / 3d reconstruction