# Some continuation on my previous posting on transformation

• April 7th, 2014, 08:18 PM
jenny_wui
Some continuation on my previous posting on transformation
Attachment 32525

Hello, this is some continuation of my previous posted topic on transformation. I have attached two pictures of rotation of different axes, i.e. X, Y and Z. Z rotation is the projection on XY plane for two different Z rotation. It shows how the length of the axis varies depending on Z rotation. Now my question is, from this information is it possible to figure out rotation of Z axis. Here as rotation is on XY plane, the length of X and Y axis does not vary, but that of Z-axis varies.

• April 8th, 2014, 03:43 AM
VictorN
Re: Some continuation on my previous posting on transformation
Quote:

Originally Posted by jenny_wui
... Z rotation is the projection on XY plane for two different Z rotation. It shows how the length of the axis varies depending on Z rotation.

Your definition of "rotation" is not clear! How would it be that "Z rotation is" something "for two different Z rotation"? :confused:
• April 8th, 2014, 04:03 AM
2kaud
Re: Some continuation on my previous posting on transformation
As this isn't specifically a c++ question, you might get more guidance if you posted to the Algorithm forum.
• April 8th, 2014, 07:07 AM
OReubens
Re: Some continuation on my previous posting on transformation
If what you mean is... from a given projected image (result) and a given input model figure out the amount of rotation along each axis:
then no this isn't unambiguously solvable for all possible input cases.

In the 3D to 2D projection phase, information is being lost.