EigenQuaternionParameterisation does not preserve unit magnitude

47 views
Skip to first unread message

dr.a.k...@gmail.com

unread,
Mar 8, 2022, 5:54:43 AM3/8/22
to Ceres Solver
Hi

As per the title, setting a 4-block as EigenQuaternionParameterisation does not preserve the magnitude of this 4-vector as 1.

Is this expected/designed behaviour, or is it a bug?

Thanks

1262871536

unread,
Mar 8, 2022, 5:54:57 AM3/8/22
to dr.a.k...@gmail.com
这是来自QQ邮箱的假期自动回复邮件。

您好,我最近正在休假中,无法亲自回复您的邮件。我将在假期结束后,尽快给您回复。

dr.a.k...@gmail.com

unread,
Mar 8, 2022, 5:59:12 AM3/8/22
to Ceres Solver
Ahh - it seems to preserve whatever magnitude was sent in. Maybe this is the expected behaviour?

I note that initialising an Eigen::Quaterniond via a rotation matrix does not guarantee a unit quaternion!

1262871536

unread,
Mar 8, 2022, 5:59:26 AM3/8/22
to dr.a.k...@gmail.com

Sameer Agarwal

unread,
Mar 8, 2022, 8:05:04 AM3/8/22
to ceres-...@googlegroups.com
That's a bug in Eigen. It should produce a unit quaternion.
Sameer 

--
You received this message because you are subscribed to the Google Groups "Ceres Solver" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ceres-solver...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ceres-solver/0e9b512a-638e-47e1-93a1-d643a2d2ee7fn%40googlegroups.com.

1262871536

unread,
Mar 8, 2022, 8:05:11 AM3/8/22
to Sameer Agarwal' via Ceres Solver <>
Reply all
Reply to author
Forward
0 new messages