added 2 commits

B1ueber2y changed the title Raise error when reconstruction.AddImage() add a non-existent image in the corresponding frame. Raise error when reconstruction.AddImage add a non-existent image in the corresponding frame.

B1ueber2y changed the title Raise error when reconstruction.AddImage add a non-existent image in the corresponding frame. Raise error when image does not exist in its associated frame.

B1ueber2y changed the title Raise error when image does not exist in its associated frame. Throw exception when image does not exist in its associated frame.

B1ueber2y changed the title Throw exception when image does not exist in its associated frame. Throw exception when image.SetFramePtr() sets a wrong frame.

Merged

B1ueber2y changed the title Throw exception when image.SetFramePtr() sets a wrong frame. Throw exception when image.SetFramePtr() sets a frame without the image listed as its data.

B1ueber2y added a commit that referenced this pull request

…ciated rig. (#3543)
As a follow up to #3540 for safety
improvement, this PR includes the following fixes:
* ``AddFrame`` should not add a frame with empty data as its never
useful in the reconstruction. NumData() is added for the frame class.
* ``Frame.SetRigPtr`` should set a rig that has all sensors in the frame
data supported.
* Conversely, when a rig pointer is available for the frame,
``Frame.AddDataId`` should add data that with the sensor available in
the associated rig.
Related tests are added.
---------
Co-authored-by: Johannes Schönberger <jsch@demuc.de>

Merged

B1ueber2y added a commit that referenced this pull request

…ame does not exist in the reconstruction. (#3546)
Follow up to #3540 and
#3543
---------
Co-authored-by: Johannes Schönberger <jsch@demuc.de>

tavislocus pushed a commit to tavislocus/colmap_6dof that referenced this pull request

…ge listed as its data. (colmap#3540)
Otherwise the reconstruction data structure is broken, and the round
trip wont write out the images.
Seems to catch issues in some of the test construction.
---------
Co-authored-by: Johannes Schönberger <jsch@demuc.de>

tavislocus pushed a commit to tavislocus/colmap_6dof that referenced this pull request

…ciated rig. (colmap#3543)
As a follow up to colmap#3540 for safety
improvement, this PR includes the following fixes:
* ``AddFrame`` should not add a frame with empty data as its never
useful in the reconstruction. NumData() is added for the frame class.
* ``Frame.SetRigPtr`` should set a rig that has all sensors in the frame
data supported.
* Conversely, when a rig pointer is available for the frame,
``Frame.AddDataId`` should add data that with the sensor available in
the associated rig.
Related tests are added.
---------
Co-authored-by: Johannes Schönberger <jsch@demuc.de>

tavislocus pushed a commit to tavislocus/colmap_6dof that referenced this pull request

ahojnnes added a commit that referenced this pull request

…ge listed as its data. (#3540)
Otherwise the reconstruction data structure is broken, and the round
trip wont write out the images.
Seems to catch issues in some of the test construction.
---------
Co-authored-by: Johannes Schönberger <jsch@demuc.de>

ahojnnes added a commit that referenced this pull request

…ciated rig. (#3543)
As a follow up to #3540 for safety
improvement, this PR includes the following fixes:
* ``AddFrame`` should not add a frame with empty data as its never
useful in the reconstruction. NumData() is added for the frame class.
* ``Frame.SetRigPtr`` should set a rig that has all sensors in the frame
data supported.
* Conversely, when a rig pointer is available for the frame,
``Frame.AddDataId`` should add data that with the sensor available in
the associated rig.
Related tests are added.
---------
Co-authored-by: Johannes Schönberger <jsch@demuc.de>

ahojnnes added a commit that referenced this pull request

…ame does not exist in the reconstruction. (#3546)
Follow up to #3540 and
#3543
---------
Co-authored-by: Johannes Schönberger <jsch@demuc.de>

Merged