Obtaining/Creating 4x4 Transformation matrix to the location of a vtkMRMLCameraNode

Development
1

I’m currently working on a project where I control the position of a vtkCameraNode in 3D view with a robotic arm. To achieve this, I’ve been placing the camera node on top of a transform like so:


camera_node.SetNodeReferenceID("transform", transform_for_the_camera.GetID())

With this method, I can successfully move the camera and retrieve the position of the camera using GetAppliedTransform(). Here is an example of the output:


print(stereo.camera_node2.GetAppliedTransform())

Output:


vtkMatrix4x4 (0x564de95445d0)
  Debug: Off
  Modified Time: 2695454
  Reference Count: 2
  Registered Events: (none)
  Elements:
    -0.0177592 0.999842 7.34635e-06 -0.000133888 
    0.00592044 9.78115e-05 0.999982 -36.4499 
    0.999825 0.0177589 -0.00592124 -12.88 
    0 0 0 1

However, I have a couple of challenges that I need help with. Can I either? :

1. Is there a way to obtain the transformation to the camera_node before SetNodeReferenceID(“transform”, transform_for_the_camera.GetID()) is used?

OR

2. Is it possible to reconstruct the AppliedTransform using available functions like GetViewUp() and GetPosition()?

To illustrate the second point, here are the outputs of these functions:


print(stereo.camera_node2.GetViewUp())  # view_up
print(stereo.camera_node2.GetPosition())  # position
print(stereo.camera_node2.GetFocalPoint())  # focal_point

Output:


(7.346345810766298e-06, 0.9999824692637288, -0.005921242373602143)  # view_up
(-0.00013388794915210472, -36.4498973704036, -12.879999999811874)  # position
(-1.2295649116714942, -86.05915036780942, 10.081280602927164)  # focal_point

In the AppliedTransform 4x4 matrix, the view_up vector matches the third column and position matches the fourth column. I attempted to reconstruct the first or second column using (focal_point - position)/norm(focal_point - position) but the results do not match the values in the matrix.

I would greatly appreciate any guidance or suggestions. Thanks in advance for your time and help!

2

I would recommend to use the Viewpoint module in SlicerIGT extension for positioning a camera with a transform. It allows you to directly set the relative transform between the driving transform and the camera.