ROI Statistics 3D yaw angle and pitch angle

[Jingjing]

Dear Icy team,

First thanks for the wonderful active contours plugin that segments my 3D cells quite accurately! 

I have another question here regarding the measurement of the segmented 3D objects - now I need to extract the unit vectors of the major axes of the (elongated) cells and use them for further calculations. In the plugin ROI Statistics I found the measurements "yaw angle", "pitch angle" and "roll angle", and could I ask whether it is pitch angle that I need to use to get the unit vector of the major axis? I originally thought so but after taking a look at a couple of objects I was not quite sure, because yaw angle actually looks more like the angle I need to extract...

I didn't find the information about these measurements in the documentation so thought I'd better confirm with you, 

Thanks a lot,

Jingjing

[Alexandre Dufour]

Hi Jingjing,

The “yaw”, “pitch” and “roll” are very recent additions, hence the substandard documentation (I will fix this ASAP). Basically they correspond to the angles around the vertical, lateral, and longitudinal axes, respectively, following this popular orientation system.

I you require the actual axes themselves, then you should have a look at the “1st axis”, “2nd axis”, and “3rd axis” descriptors (they are all normalised to unit-length). The only trouble (for now) is that the entire vector is displayed in a specific text field looking like "(0.0, 1.0, 0.0)”, so you’d have to parse the numbers inside the brackets to get to the actual values.

Note: all of these measures are available in both the “ROI Statistics” and Icy’s very own “ROI” panel.

Alexandre

--
You received this message because you are subscribed to the Google Groups "Icy imaging" group.
To unsubscribe from this group and stop receiving emails from it, send an email to icy-software...@googlegroups.com.
To post to this group, send email to icy-so...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/icy-software/b7e10bed-e8f1-4262-b3ad-add8db4f3eeb%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Jingjing]

Dear Alexandre,

Thanks for your quick reply! I had a go with the 1st axis vector and found the angle between the object axis and the z axis (of the entire stack) match the pitch angle, is it correct? And don't worry about the character-form numbers, I can parse them into values. 

As a follow-up, could I ask what algorithm is used to find the "best fitting ellipsoid" for the three axes? I need to go through quite a few cells to validate my segmentation result so would love to learn a bit more how the result is generated.

Thanks again,

Jingjing

[Alexandre Dufour]

Dear Jingjing,

The ellipsoid fitting is based on the least-squares algorithm available here: http://www.mathworks.com/matlabcentral/fileexchange/24693-ellipsoid-fit

Alexandre

To view this discussion on the web visit https://groups.google.com/d/msgid/icy-software/f4ace80e-9fb3-49fa-b5ed-9d2a8fe33d47%40googlegroups.com.