search

rpng / open_vins

An open source platform for visual-inertial navigation research.
https://docs.openvins.com
GNU General Public License v3.0
2.06k stars 616 forks source link

A question for predict_mean_discrete function in Propagator. #307

Closed ilseojeon closed 1 year ago

ilseojeon commented 1 year ago

In predict_mean_discrete function in Propagator class,

I've noticed you commented // new_q = rot_2_quat(exp_so3(-w_hat*dt)*R_Gtoi);

Is there a particular reason you didn't use SO3 to predict the orientation?

goldbattle commented 1 year ago

I believe they gave the same performance and both make the same constant measurement assumption. The quaternion orientation is consistent with the original MSCKF paper, so I think we just selected that one. Hope this helps.

On Thu, Feb 16, 2023 at 6:17 AM ilseojeon @.***> wrote:

I've noticed you commented // new_q = rot_2_quat(exp_so3(-w_hatdt)R_Gtoi);

Is there a particular reason you didn't use SO3 to predict the orientation?

— Reply to this email directly, view it on GitHub https://github.com/rpng/open_vins/issues/307, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAQ6TYQJG3BMNNPFUMSHQYLWXYEDLANCNFSM6AAAAAAU6BIKNA . You are receiving this because you are subscribed to this thread.Message ID: @.***>