[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vmg / vr / vrpg / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Search] [Mobile] [Home]
Settings Mobile Home
/3/ - 3DCG

Thread archived.
You cannot reply anymore.

File: file.png (669 KB, 855x557)
669 KB
669 KB PNG
I want to do some photogrammetry to 3D print via Shapeways with the sandstone option they offer, so the print would be in colour. I don't know much about 3D stuff but both methods i've seen seem pretty retard proof, i just i'm just seeking some guidance in wich one to use:

Should i go with the Skanect + Xbox Kinect method?

Or the Meshroom+Camera method?

Im mostly interested in wich one would end up with the most details even if it ends up being harder. Please and thank you guys.
File: Untitled.png (177 KB, 299x237)
177 KB
177 KB PNG
Meshroom+Camera method
Meshroom + Camera, it will take a bit longer but you'll get much better results.
Thanks guys, i've been making some test runs with this method and been having some nice results just got to tweak some of my steps so it works perfectly this will work for what i want to do.
>photogrammetry to 3D print
That's not all that easy to get good results from.
Particularly if you're trying to get small details off existing miniatures. (for that you need photography skills, an expensive DSLR camera, an even more expensive lens and dealing with Meshroom/3DF Zephyr/etc)
I use Metashape PE and a Leica S3 at work
>Leica S3

Sounds like a dream. What do you make models of?
File: 3.jpg (949 KB, 888x888)
949 KB
949 KB JPG
Im pretty adept at photography so that part is no issue, i dont think at least, im taking all the shots on raw to keep consistency image by image and shooting in manual with a prime lense to avoid warping. I'm actually trying to scan people and have been fine tuning the process but still having issues. So far the only guides ive found for this specific thing has been with static architecture or a person T posing and doesnt really go in detail on the early parts, but i want to scan rigid specific poses.

To mitigate the movement of the subject ive been scanning the person with a second photographer so more pictures can be taken quicker because holding still for more than 7 minutes seems to be unreasonable for the models ive tried it with so far. Weve been doing it at golden hour on a rooftop to have both enough space to orbit around the subject and have even lighting, we end up taking 140+ shots each session.

Of course this silly DIY approach doesnt compare to properly doing so with a dedicated room with a sphere of cameras but Meshroom delivers a rough enough approximation that i can then sculpt using blender into the person using the reference pictures.

I think right now the biggest obstacle is the texturing, wich always ends up being 4+ png files, sometimes it includes a big chunk of say the face but its all holed up. I can take that and complete that bit on photoshop matching the picture with that got wrecked, the problem is that its mostly comparmentalized in this incomprehensible blizzard like way. I think if i can find a way to redo the uv mapping for me to then fix up the individual textures in 2d matching and improving in photoshop then id be done. But like i said im very new to 3d stuff i dont know if this is the proper way to handle all of this or if im talking nonsense.

For what its worth mentioning my pipeline so far has been: Taking the pictures >Lightroom>Meshroom>Blender.
Hello OP.
I'm at a private instituion and we have conducted various photogammetry captures.
Although we have Einscan tools we have been able to get succesfully strong models with great details with a newer version of Iphone. The process is as follows.

1. Good Lighting: All around even lighting, Avoid hard contrast on the object

2. A method of rotating the object while we performed the capture ( lazy susan works or any rotating thing you can dfind in your house)

3. Film a high res video with your camera of choice ( Iphone or greater)

4. Make sure to capture the object in elevation view and slowly rotate to capture it from top view ( aim for about 40 seconds of total film time)

4. Obtain the mp4 file and run it through VLC player.

5. Enable Film splitting in VLC player. It will give you individual frames in jpegs or file format of your choice.

6. Run the images through Autodesk Recap. If you have a university email get the free student version it will work just fine.

Recap will then reconstruct the object into a high res mesh with textures ingrained as a separate layer, plus it will provide a texture map from said scan.

Any separate questions feel free to ask.
DIdnt Read this at first,

You can still achieve decent quality scans as long as you have adequate lighting.
All the people we have scanned have been with the Einscan tools.
That will happen if you use raws, convert them to tif or PNG before meshing
File: Photogrammetry.jpg (112 KB, 929x1033)
112 KB
112 KB JPG
if you want to try and scan miniatures i strongly recommend looking at this site, it's a museum that photoscans miniatures and has a quite in depth guide on how to do it.

I'm pretty new to 3d modeling, what do you guys use to touch up the models, and how many photos do you take for each model?

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.