Main Content


(To be removed) Returns camera poses associated to views

The viewSet object and its poses function will be removed in a future release. Use the imageviewset object and its poses function instead. For more information, see Compatibility Considerations.



cameraPoses = poses(vSet) returns the camera poses that correspond to the views contained in the input viewSet object, vSet.

cameraPoses = poses(vSet,viewIds) returns the camera poses that correspond to a subset of views specified by the vector viewIds.


collapse all

Create an empty viewSet object.

vSet = viewSet;

Add views to the object.

vSet = addView(vSet,1,'Orientation',eye(3),'Location',[0,0,0]);
vSet = addView(vSet,2,'Orientation',eye(3),'Location',[1,0,0]);

Retrieve the absolute camera poses.

camPoses = poses(vSet);

Input Arguments

collapse all

viewSet object.

View IDs, specified as an integer scalar for a single view, or as a vector of integers for a set of views.

Output Arguments

collapse all

Camera pose information, returned as a three-column table. The table contains columns for ViewId, Orientation, and Location. The view IDs correspond to the IDs in the viewSet object. The orientations are specified as 3-by-3 rotation matrices and locations are specified as three-element vectors. You can pass the cameraPoses table to the triangulateMultiview and the bundleAdjustment functions.

Version History

Introduced in R2016a

expand all

Not recommended starting in R2020a