search

MIT-SPARK / TEASER-plusplus

A fast and robust point cloud registration library
MIT License
1.69k stars 333 forks source link

[BUG] scale is 0. #163

Closed 2848385707 closed 2 months ago

2848385707 commented 1 year ago

Describe the bug A clear and concise description of what the bug is.

Have you run the unit tests?

Installed Dependencies Please report the versions of all dependencies you installed, as well as your OS.

To Reproduce Please post the exact steps you executed to reproduce the bug.

Additional context Add any other context about the problem here. I used my own data set and used scale estimation. I entered my own feature matching pairs for teaser++, and finally the scale estimated by teaser++ was 0 and the rotation matrix was np.ones((1,1,1)). I wanted to know if the feature matching pairs I used were not correct enough. Thank you in advance. VOGC_5`B9}{QM( `143SE7O

jingnanshi commented 1 year ago

It seems like the rotation noise bound set is infinite, which should not occur. Did you explicitly set the parameters for the solver when you use the Python API?

LimHyungTae commented 2 months ago

If there are no further questions, we will close the issue. Thx!