You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Thanks for this amazing work leading the trend for monocular SLAM!
I found two possible bugs in the part that normalizing the quaternion after EKF. Please check and correct me if I was wrong.
the Jacobian for dq_norm_by_dq. Specifically, in the function dqi_by_dqi() and dqi_by_dqj() in motion_model.cpp, the "qq" should all be "sqrt(qq)", according to my derivation
the quaternion is not successfully normalized in function func_xvnorm_and_dxvnorm_by_dxv in motion_model.cpp. After this function , the quaternion in state vector is still not an unit quaternion. In function "func_xvnorm_and_dxvnorm_by_dxv", Tempqb should be divided by its norm before passing its value to "xvnorm"
When I run the Scenlib2, the quaternion norm is increasing by the time, this value is close to 1 during a short time, so the performance is not infected significantly. But when running in a long period, this could be a serious problem.
The text was updated successfully, but these errors were encountered:
Thanks for this amazing work leading the trend for monocular SLAM!
I found two possible bugs in the part that normalizing the quaternion after EKF. Please check and correct me if I was wrong.
Tempqb should be divided by its norm before passing its value to "xvnorm"
When I run the Scenlib2, the quaternion norm is increasing by the time, this value is close to 1 during a short time, so the performance is not infected significantly. But when running in a long period, this could be a serious problem.
The text was updated successfully, but these errors were encountered: