[HN Gopher] 3D Scanning Systems That Can Render a Complete, Edit...
       ___________________________________________________________________
        
       3D Scanning Systems That Can Render a Complete, Editable 3D Model
       in Minutes
        
       Author : simonebrunozzi
       Score  : 31 points
       Date   : 2021-03-21 07:55 UTC (1 days ago)
        
 (HTM) web link (lynceans.org)
 (TXT) w3m dump (lynceans.org)
        
       | holoduke wrote:
       | Every once in a while I start looking for a great affordable 3d
       | scanner. Is there something like the equivalent of the prusa 3d
       | printer for 3d scanning?
        
         | errantspark wrote:
         | No because 3D scanning is really, really hard. Much much more
         | complex than 3d printing. There's no fool-proof way of
         | determining the position and normal of a surface point other
         | than to touch it. All EM based techniques must contend with the
         | wide gamut of different interactions common materials will have
         | with a given wavelength. Not to mention occlusion.
        
         | CobrastanJorji wrote:
         | Good question. There are several 3D scanners in that price
         | range, but I'm curious as to the difference between
         | photogrammetry with a phone, something like the $800 SOL 3D,
         | and the $10k+ handheld options.
         | 
         | The ~$1k 3D scanners out there advertise an accuracy of
         | something like "up to 0.1mm," which puts them in the ballpark
         | of the resolution a basic Prusa's likely to be able to
         | reproduce, but I'm really unclear on how true that is in
         | practice.
        
           | errantspark wrote:
           | There's likely some actual Nyquist-Shannon ish corollary that
           | will tell you how much higher your scanning res has to be
           | than your printing res, but 2x is probably a safe bet. Not to
           | mention "accuracy" tells an imprecise story. Does it mean
           | that the best case scenario every point is at most 0.1mm from
           | where it should be? That doesn't mean the same thing as
           | "accurately resolves features larger than 0.1mm3".
        
         | valine wrote:
         | Any iPhone with FaceID is actually a very capable as a 3D
         | scanner. I recommend heges, it has support for Lidar scanning
         | as well (if you have an iPhone 12 Pro or a new iPad Pro).
         | 
         | https://hege.sh/
        
       | nobbis wrote:
       | Cubify was a rebranded Structure Sensor, which was a Primesense
       | camera, which was the original Kinect (that Apple bought in
       | 2013.)
       | 
       | It used structured light: project a static pattern of infrared
       | dots and then measure parallax using an infrared camera a couple
       | inches from the projector to calculate depth. It worked well, but
       | can't be miniaturized as it relies on a large offset between
       | projector and camera. Orbbec make similar cameras today, e.g.
       | https://orbbec3d.com/product-astra-pro
       | 
       | Modern depth cameras, like the LiDAR Scanner in iPhones and
       | iPads, use a different technique: time-of-flight (ToF), which
       | shoots laser pulses and measures the time taken to reflect back
       | to infer depth. Apple leads the field here, because they're the
       | first to integrate a "direct" ToF sensor into mobile devices,
       | which has a number of advantages over the "continuous wave" ToF
       | found in Android phones.
        
         | burntoutfire wrote:
         | Do you know if purely vision based techniques have any
         | advantages for inferring depth in these scenarios (close range
         | object scanning) over lidars?
        
           | nobbis wrote:
           | Yes, different failure cases (purely vision can't tell the
           | depth of a plain white wall) but there are orders of
           | magnitude more pixels in regular cameras. Apple's LiDAR
           | produces a 256x192 depth map, but its RGB cameras run at
           | 1920x1440. You get much more detail, but passive stereo is
           | orders of magnitude slower (ToF sensors are 60 Hz,
           | photogrammetry takes minutes.)
           | 
           | They complement each other well, which is the basis of our
           | approach at Abound: combine the detail of multi-view stereo
           | for short-range with the robustness of ToF out to 5m.
        
       | slobiwan wrote:
       | This is a pointer to a review from 2015 for a product that's no
       | longer commercially available. Seems like perhaps the technology
       | wasn't as mature or capable as the review seems to indicate, if
       | it didn't make it.
        
         | vertis wrote:
         | Having owned and used the device I can attest to that. It was
         | hard to get it to work reliably, and it certainly couldn't
         | handle either small finely detailed items or bigger spaces like
         | whole rooms.
         | 
         | It would do part of a room fine and then lose tracking and
         | never be able to re-establish tracking and you would have to
         | start all over again.
        
           | errantspark wrote:
           | The V2 (Structure Core) has a built in IMU and is much better
           | at keeping tracking. The quality of the USB link between the
           | sensor and the computer also makes a big difference. (not an
           | expert in this particular area) but iirc USB is like TCP in
           | that if a packet is missing or corrupted it'll ask for it to
           | be resent, it seems like the Structure sensors are quite
           | sensitive to this and missing more then a few frames leads to
           | alignment issues. That being said I wouldn't call the
           | experience user friendly on any level, their software is
           | really really bad too, I've had to write some tools to modify
           | their save files so that I can rotate the reconstruction area
           | because that's not a feature of the software. Definitely
           | strongly dis-recommend buying one.
        
       | ffffwe3rq352y3 wrote:
       | This could be a big deal for games Games Workshop and Warhammer
       | like games in general. I don't think this could copy model
       | because they are so small and detailed but the technology is
       | advancing quickly! Why bother buying the models when I can 3d
       | print them OR port the 3d file into TTS and play online?
        
         | errantspark wrote:
         | Because buying the models even at GW prices is likely cheaper
         | than the time and effort necessary to do a good job of printing
         | them. Model piracy does occur but it takes the from of Chinese
         | recasters making molds directly from the model sprues. That all
         | being said there _is_ a market for this and my guess is the BOM
         | cost of the hardware necessary to make a 3D scanner capable of
         | resolving the detail of a tabletop mini is in the $10-$30 range
         | right now. No way there 's enough market to foot the bill of SW
         | dev tho.
        
           | ffffwe3rq352y3 wrote:
           | Yeah with the current tech it isn't really worth it to print
           | stuff unless its like an entire army and you would be ok not
           | using GW models. However, in the future I'm betting this will
           | change and It will be much cheaper to print the models at
           | home or have someone you know do it!
        
         | simonebrunozzi wrote:
         | I wish I had a system like this when I was a kid, and spent a
         | the equivalent of a few hundred euros over the course of a few
         | years to buy just a few of these miniatures. These prices were
         | simply outrageous, especially outside the UK.
        
         | ilaksh wrote:
         | You can search for 3D scan in your app store for current
         | products and then try it.
        
       | belval wrote:
       | This is based on structured/coded light and as such is basically
       | useless for most surfaces that will not reflect light (think most
       | black things) or very reflective things (think metal). I have an
       | SR300 which was intended for face scans and it works very nicely
       | for humans but I tried to scan other surfaces for a school
       | project[1] and the limitations became obvious very fast.
       | 
       | [1] https://github.com/Belval/Scanner3D
        
       | valine wrote:
       | These scans don't look any higher quality than what you can get
       | with FaceId or lidar on the iPad Pro. It's a product from 2015 so
       | I guess it could have made sense back then.
        
       ___________________________________________________________________
       (page generated 2021-03-22 23:02 UTC)