Re-use Alignment
Answered
What is the procedure for this? I have a good scan and want to add a few more photos
-
Wishgranter will certainly answer you, but on what I see when I add pics to a successful aligment, the system speed up the aligment process, reusing datas saved in the cache system.
One thing important for me was to set the cache directory to a specifical hd (used only for that).
But when I add "macro" pictures to a succesful aligment, I encounter no problem at all and it is fast... -
Hi Wishgranter . I have a few
1: Turntable scan of an object eg statue. Main images 80 Percent aligned and i took some extra photos to "patch up the occluded parts". Most of the time i have to use CPs for the next set of images since i reposition the object to shoot the feet.
2: Head scan, where i want to lock parts of the body from realignment. eg face and
3: Full body. Again some parts of the alignment I want to lock or reuse. Add more photos and CPs
4: Aerial, and large structures like buildings. I may have 2 or 3 drones covering a structure.
I usually add images incremental picking the wide coverage shots to get the main structure then add close ups for details.
Usually the alignment is good up to a certain point where it really slows down. -
I also want to know how I can re-use alignment. I did 4 scans in one day and the cameras remained in the same position. I got nice alignment produced tom CapturingReality for one set and I've like to save that and apply it to the other sets. The cameras from every set have matching filenames (serial number of the camera). -
1: Turntable scan of an object eg statue. Main images 80 Percent aligned and i took some extra photos to "patch up the occluded parts". Most of the time i have to use CPs for the next set of images since i reposition the object to shoot the feet.
RC stores "detected points" in the CACHE folder, so when you add some images, it will use already detected points ( if you have not changed alignment settings significantly ) and detect points in newly added images.2: Head scan, where i want to lock parts of the body from realignment. eg face and
Is it a fixed camera head scanner or do you use a "turntable" scenario ?
If it is fixed and the cameras have not moved, just the subject is changed, then just select all cameras and export them as metadata (XMPs) with LOCKED settings, the XMPs are saved next to the images used. You need to COPY them into the next data which you need to align, then just load the images into a new scene and the XMPs ( as you copied them next to images ) will be used to get properly aligned data.
The XMPs transfer information about camera positions etc., so RC will use them fully automatically when they are placed next to the images loaded in RC.3: Full body. Again some parts of the alignment I want to lock or reuse. Add more photos and CPs
Same as in 2.4: Aerial, and large structures like buildings. I may have 2 or 3 drones covering a structure.
Same as in 2.I usually add images incremental picking the wide coverage shots to get the main structure then add close ups for details.
Usually the alignment is good up to a certain point where it really slows down.
I‘m not sure what you mean by the slowdown issue... -
I also want to know how I can re-use alignment. I did 4 scans in one day and the cameras remained in the same position. I got nice alignment produced tom CapturingReality for one set and I've like to save that and apply it to the other sets. The cameras from every set have matching filenames (serial number of the camera).
If you have good alignment, then select all cameras, export them as metadata (XMP) - in the ALIGNMENT ribbon with LOCKED settings. Then just COPY the XMPs to another dataset and load the data ( images ) into RC and hit align.
Please sign in to leave a comment.
Comments
14 comments