[MINC-users] MINC 2.0 coordinate system

[Soren Christensen]

Thanks - that clears it up for me.

Soren


On Thu, Dec 20, 2012 at 5:01 AM, Peter Neelin <peter.neelin at gmail.com>wrote:

> On Dec 20, 2012 1:34 AM, "Soren Christensen" <sorench at gmail.com> wrote:
>
> >  In the MINC 2.0 coordinate system section of
> >
>
> http://en.wikibooks.org/wiki/MINC/Reference/MINC2.0_File_Format_Reference#MINC_2.0_Structural_Attributes
> > there is mention of "Startx, Starty and Startz" - just after the line
> > "and the origin (0,0,0) in voxel coordinates is located at world
> > coordinates (ox,oy,oz) as given by:"
> >
> > I got a bit confused here. The Start values for x y and z that we see
> > using mincinfo are the world coordinates (referred to as "o") in that
> > section right?
>
> Nope. The start values are as described in that section (assuming that the
> matrix order is fixed). Only when the direction cosines are the non-rotated
> ([1 0 0], [0 1 0], [0 0 1]) do the start values match the origin.
>
> This convention was chosen so that applications that ignore direction
> cosines will have the correct origin in the rotated frame (start and step
> are expressed along the dimension and are self-consistent). Applications
> that want to use the full coordinate information can do the right thing -
> it is not hard, and you will almost certainly already have code doing
> matrix math, so what is a little more?
>
> > Isn't the order of these matrix multiplications the wrong way around by
> the way?
>
> Yup. That should be fixed.
>
> Peter
> --
> Peter Neelin
> peter.neelin at gmail.com
> _______________________________________________
> MINC-users at bic.mni.mcgill.ca
> http://www.bic.mni.mcgill.ca/mailman/listinfo/minc-users
>