Tried the uploading file function but cannot see them in the post.
And the other one
Hi Hennie van der Merwe
can upload it to say Dropbox and put link here to it ?
Dropbox
screen.png (file://MININT-I7ITI9M/Users/Administrat … screen.png)
screen 2.png (file://MININT-I7ITI9M/Users/Administrat … en%202.png)
Thanks
Hi Hennie van der Merwe
the links not working its refering to file direct on your disk
if have Dropbox, right click on the file and should get its drobo link…
like this one https://dl.dropboxusercontent.com/u/150 … ok_007.jpg
Now you see why I battle!
https://www.dropbox.com/s/7vzynv8dmaj07 … 2.png?dl=0
https://www.dropbox.com/s/zjua0p1f84ye3 … n.png?dl=0
Hi Hennie
the main problem is you need capture it a bit different way. Putting it on table is not the best solution…
try to put it so can shoot around whole project, sometimes its tricky but take a look on some screens here how to shoot
Thanks but it does not really solve the problem as there are lots of detail at the bottom of the object that will be masked by whatever it is suspended on. Also not a very practical solution as one iether needs a ladder for the top views or lie on your back for the bottom views - not very elegant solution.
Surely there must be a way to merge?
Hi Hennie
Understand you, but if need solve it ELEGANT way, find a pose so it can be placed on small “support” and put there something like glue (so can remove the subject easy ) you trying to scann sides, try scann it as whole object in 360 deg view…
And then capture from sides like the bonsai in say 2-3 elevations
it would be possible to make from 2sides but it need a lot of experience to put it later together as need sub mm precision on scale.
Thanks for that so my understanding is that the Merge function only works easily when more than one model was created from the same set of pctures due to not enough overlap, etc?
With reference to the glue bit I just need to make sure that all pics must be taken without moving the object? your statement could be interpreted as removing and take pics of the sides.
Looking at your screenshots it looks like more than one model was created even with all the overlap in the pics.
how does one then merge these models?
Pardon the stupid questions but I will get there eventually. LOL
Hi Hennie
the components are for putting stuff together but NOT like you want to use it. You get issue sin recon as its NOT proper photogrametry workflow…
WIshgranter I hate to admit but you now have me totally confused because your comment does not appear to answer any of my questions. Apologies once more but I did my best to find answers in the tutorials and here but only reverted to you as a last resort.
Thanks for that so my understanding is that the Merge function only works easily when more than one model was created from the same set of pctures due to not enough overlap, etc?
no its intended for large projects like this one https://www.dropbox.com/s/ym9a0ogb7ch6t … 4.png?dl=0
With reference to the glue bit I just need to make sure that all pics must be taken without moving the object? your statement could be interpreted as removing and take pics of the sides.
No, put the subject on small platform and glue it so the glued part is just small part of the subject “hidden” say bottom part of the shell
Looking at your screenshots it looks like more than one model was created even with all the overlap in the pics.
how does one then merge these models?
= there can be MULTIPLE reconstructed models under component ( look on the castle screenshot there are 2 models )
Pardon the stupid questions but I will get there eventually. LOL
No question is stupid 
look for some tutorial on photogrammetry to get the idea hwo it works… it will pay back in better understanding the whole process…
Sorrymore questions: but thanks for the detailed answers above.
will a turntable work?
can I tilt the object on a turntable to get different views or does the objects change in orientation with respect to the background create a problem?
thanks
Hi Hennie
will a turntable work?
yes its best way for this sort of stuff BUT looking on it as its SHINY and “feature-less” so it can be tricky
highly recommend that you start with something that WILL work, not shiny, not featureles like cup from porcelain without any prints on it
try something like pine cone, small statue or etc so get the proper workflow on working subject and then try different stuff to get understand what work and what not… its the safe way…
can I tilt the object on a turntable to get different views or does the objects change in orientation with respect to the background create a problem?
best if have clean background, say white background so its not detect too much points on the background
Much thanks again I now can make some progress but one last question (for now)
If object fills most of view and is close to camera with background out of focus will that help?
Also just need to make sure that I interpret your answer above correctly - can I maintain one camera position (tripod) whilst tilting object on turntable for more views? Obviously one will adjust height of camera on tripod for more angles.
Hi
Much thanks again I now can make some progress but one last question (for now)
OK
If object fills most of view and is close to camera with background out of focus will that help?
= YES
Also just need to make sure that I interpret your answer above correctly - can I maintain one camera position (tripod) whilst tilting object on turntable for more views? Obviously one will adjust height of camera on tripod for more angles.
its OK so have been done even the Bonsai scan.
Hey guys,
I have similar issue with turntable project.
The program found 2 components. Each component has 100 images with different camera angle. So I have front 360 (perpendicular to object) and a little bit from above object to see the top surface.
The questions is:
- how to merge this two components with different angles, or export one model of this two components.
- I watched tutorials at least 3 times with a deer
but I can not understand how the control points works. I am not sure my controls points was applied. To test this I choose two images where I wish to define the control points. The first one is in “component 0”, the second image is out of components and there were not any control points.
I defined 6 control points in each image. What should I do next to see the second image in component?
Here is on the video: https://youtu.be/d8naLEtLqDY?t=72 shown selected control point in 3D view, but I have control points only in 2D view.
Hi Dmitry
take a look on this video how i work with CPs
https://www.youtube.com/watch?v=d8naLEtLqDY
and let me know if can solve it or not