add images to an existing RC Component / Camera pose
Answered
Hello everybody,
i´ve got a question regarding an enhancement of a component / camera pose with some new additiv images without loosing the defind groundplane and the position, rotation and scaling of the existing cameras.
The existing alignment should be enhanced with some more, new pictures for texturing.
Is it possible without a complete new alignment? I don´t wanna loose the srt-orientation of the component, because it was already exported to Maya and get a lot of work there.
Is it possible to "grab" or lock some existing srt positions data from existing parts (ground plane rotation / scale) and use them for a workaround?
Of course i would prefer a smart direct way in Reality Capture?
Best,
zed
i´ve got a question regarding an enhancement of a component / camera pose with some new additiv images without loosing the defind groundplane and the position, rotation and scaling of the existing cameras.
The existing alignment should be enhanced with some more, new pictures for texturing.
Is it possible without a complete new alignment? I don´t wanna loose the srt-orientation of the component, because it was already exported to Maya and get a lot of work there.
Is it possible to "grab" or lock some existing srt positions data from existing parts (ground plane rotation / scale) and use them for a workaround?
Of course i would prefer a smart direct way in Reality Capture?
Best,
zed
-
Hi Uwe Zahn
Try this workflow:
go to the ALIGNMENT ribbon and find SELECT ALL ( cameras ), with that you select all existing cameras, then use the METADATA (XMP) tool and export XMPs with settings in the screenshot004.png
RC will export the XMP metadata next to images with all the camera parameters ( position, orientation, scale ) that are LOCKED.
When it finishes the export ( few seconds ), restart RC and load the same scene again..
The next step is to load the additional images in the particular scene and you need to ALIGN them. The XMPs will keep the same position of the cameras and scene..
Once the ALIGNMENT is finished, you can try to import the model ( and check if it is imported to the same position ) and use texturing.... and voila - it should be solved...-
004.png
-
-
Hello Wishgater,
thank you for your response.
I´ve tried this workflow but the result was not as expected.
The ground plane rotation get lost after realingment with the new images / cameras. For this test, i only used four "new" cameras. They are aligned good in relationship to the locked, "old" cameras, but the rotation of the grond plane is different / wrong.
Here two screenshots of the alignments. The first one shows rectangular orientated and locked component. The scond shows the result after RC restart, aligning the (xmp-) locked cameras plus the four new cams. The ground is rotated round about 15 degree.RC_Alignment.jpgRC_ReAlignment.jpg
I hope you have an idea how I can solve this problem.
Regards,
Uwe-
437RC_Alignment.jpghttps://www.capturingreality.com/forum/download/file.php?id=4372097281imageFile jpgRC_Alignment.jpg
-
438RC_ReAlignment.jpghttps://www.capturingreality.com/forum/download/file.php?id=4382097281imageFile jpgRC_ReAlignment.jpg
-
Please sign in to leave a comment.
Comments
6 comments