In a Way, 3D Scanning Is over a Century Old | Hacker News Reader

In a Way, 3D Scanning Is over a Century Old

hackaday.com

No mention of photogrammetry, or its pioneer Meydenbauer in 1858?

http://www.theulegium.de/fileadmin/user_upload/Texte/Meydenb...

Great share. I also assumed the article was about photogrammetry, was surprised at the contents, and was unaware of the significance of Meydenbauer. I wonder whether he would side with the architects or the software world on the desirability of our modern use of the title 'architect' in the field of computing... I suppose this is less of a debate in German, since it's probably ManyWordsTogetherArchitect and thus clearly defined.

Fascinating! Thank you

It is a good reminder that today we are not more clever with time. We come here only with more tools, easier tools and more time to think...

Also a reminder that the 3D scanning is still unchanged after 1 century. Odd...

- But why ? 1) Because it was the good solution ? Maybe....

2) Because no one has had the new-thing that will level up the way we do 3D scanning ? I hope so

> Also a reminder that the 3D scanning is still unchanged after 1 century.

That is simply false. We have interactive games based on cheap costumer 3d scanners. [1] We have 3d scanners in our pockets. [2] We have 3d scanners which can scan millions of points per second from the environment for hundreds of meters of distance. [3]

But yeah, if we ignore all the things which changed, then it turns out everything is "still unchanged".

1: https://en.wikipedia.org/wiki/Kinect 2: https://www.techradar.com/news/what-is-a-lidar-scanner-the-i... 3: https://www.mapix.com/wp-content/uploads/2018/07/63-9378_Rev...

I assumed the OP was talking more about the math, which is also not true. The method mentioned in the article only takes into account the silhouette of the target. In my experience, that's where a lot of the information comes from, but to get that fine detail most contemporary methods incorporate a lot of information about the target itself like keypoints and lighting interactions.

Our newest product is a very advanced 3D Scanner[1]. Photogrammetry, structured light scanning, phase shift laser radar, and now non-contact via an interferometry. 3D Scanning is alive and well with various methods.

[1]https://apimetrology.com/ladar/

I think it’s also important to note that these said tools are more readily available to whoever wants to use them. Hard to say because there’s not much information about François Willème but I’m guessing he’s relatively wealthy to be able to pull something off like this.

>Because no one has had the new-thing that will level up the way we do 3D scanning ?

There isn't one way to do 3d scanning. Unless you mean any method that involves taking 1 or more pictures is the same method.

Hackaday isn't the worst "repackager", but I always like to go to the original source for the article: https://www.messynessychic.com/2022/09/28/more-than-100-year...

Vaguely relatedly, the "Cast Court" in the Victoria & Albert museum in London is full of 19th-century models of various famous bits of architecture and sculpture, mostly done by plaster cast:

https://www.vam.ac.uk/collections/cast-collection

It's a really fun visit as a progenitor of modern enthusiasm for reproductions of famous objects and monuments.

My old director made a 3d scan of Obama when he was at the Smithsonian. He keeps a 3d printed Obama bust on his desk.

"3d scanning is as old as humanity" - The Roman emperor Gluteus Maximus

Yes, The target ( of Scanning 3D objects).

But not the solution witch I bet, has had many differents solutions.

This one solution ( 'slices od 2D'), one century old, still used in 3D laser scanning and printing, is poised to have a more robust and fine grained one.

Ancient egyptians had 2d pantographs attached to linked turntables, enabling them to point in 3d.

It's possible, and one candidate explanation for how they were able to produce multiple identical immense statues at widely separated locations, but as far as I know nobody has found such a pantograph or turntable, nor drawings of it or writings about it, nor any other kind of evidence that this was the method they used.

Great quick read, tldr they took a bunch of bullet time photos in a circle around someone, made a bunch of vertical wood cutouts of the silhouettes and then wedged them together vertically / radially and filled the gaps with clay. The freaky picture helps.

Speaking of this does ANYONE have a suggestion for a 3D scanning service / app that could capture reasonable detail on a warhammer / hotwheel sized model?

I've asked around and gotten the "don't bother just remake it in blender" response pretty consistently... but I don't wana' (ie its not realistic). Somebody please help if you can.

Or I guess I could get a bunch of cameras and sheets of wood...

Scaniverse (iOS app) works pretty well on figurines and small models, for example:

- https://sketchfab.com/3d-models/kitzreh20-a37b3f78b38048d286...

- https://sketchfab.com/3d-models/taurenscaniverse-c78b86cfe9c...

Mail it to me and I will with 9D LADAR. The point cloud data will include color data on the points. Fill out a contact form on our website asking for Robert.

https://apimetrology.com/ladar/

>does ANYONE have a suggestion for a 3D scanning service / app that could capture reasonable detail on a warhammer / hotwheel sized model?

Someone sent me this a while back, quick tutorial using Meshroom and an ordinary camera;

https://peterfalkingham.com/2019/01/16/small-object-photogra...

I imagine for small objects you'll want to set the camera further back and zoom in rather than putting it close up, it'll probably be easier for the software to interpret (for the same reason that putting the camera close to your face makes your nose look huge).

Reply on news.ycombinator.com