BUG: Fix bounds computation in vtkMultiVolume

This line set the homogeneous IJK coordinates incorrectly:

pointsDataCoords.emplace_back(bounds[1], bounds[3], bounds[5], 0.);

The problem is that the last component is set to 0 instead of 1 (the coordinates specify a direction vector and not a position). When this vector is multiplied by the homogeneous transformation matrix, the position is incorrect, because the translation component of the matrix is not taken into account.

fixes https://github.com/Slicer/Slicer/issues/8301

fixes #19447 (closed)

added area:rendering label

Warnings:

• please consider adding a changelog entry in a file ending with .md in Documentation/release/dev.

The warnings do not need to be fixed, but it is recommended to do so.

+1 Thanks for the merge request @dzenanz. I've tested this change and it fixes the incorrect bound computation issue in 3D Slicer.

Thank you for the fix Andras!

requested review from @sankhesh

requested review from @cory.quammen and removed review request for @sankhesh

assigned to @sankhesh

Do: test

+2

approved this merge request

Are the two CI failures a problem? They seem unrelated to these changes.

Not a problem with this MR. Please go ahead and merge.

Do: merge

mentioned in commit 1cc216ee

merged manually

mentioned in merge request paraview/paraview!7216 (merged)

mentioned in commit jean.fechter/paraview@09a59c29

mentioned in commit jean.fechter/paraview@6ea2a210

mentioned in commit jean.fechter/paraview@04eaf193

mentioned in commit jean.fechter/paraview@2dcd25f9

mentioned in commit jean.fechter/paraview@4bdd0421

mentioned in merge request !12002 (merged)