Re: [eigen] Potential unsupported module: Lie Groups

[ Thread Index | Date Index | More Archives ]

On 07.05.2012 14:19, Hauke Strasdat wrote:
I'm afraid you missed a singularity for log of the negative unit quaternion
(w=-1, v=0).
Oh, yes, thanks for that! I checked for singularities using
"exp(log(x)) = x" tests,
but thereby I ignored the quaternion double cover. I fixed it and
updated my tests.

Still not working for w=-1+eps, v!=0. Test with a small v, and make sure that log(-exp(v))==v.

(An atan2-based implementation seems to be a good strategy
for non-unit quaternions. There is also one in LGSM.)

As pointed out in our paper, you must use atan not atan2 (due to the +/- double cover):, Section 3.6.4


Dipl.-Inf. Christoph Hertzberg
Cartesium 0.049
Universität Bremen
Enrique-Schmidt-Straße 5
28359 Bremen

Tel: +49 (421) 218-64252

Mail converted by MHonArc 2.6.19+