3D Scanner inside the printer

[th3o 0]

I'm thinking of a 3D Scanner inside the printer that can be used in the Ultimaker Original and Ultimaker 2.

Something like David's Scanner http://www.david-3d.com/

Things needed:

- turntable with a stepper motor that can be placed on the bed

- bracket that holds a camera and a laser

- software or modified firmware to lift the bed step by step while scanning

(- background with dots)

Of course the turntable and bracket has to be quickly removable when you want to print

What do you think of the pro's and con's of a setup like that?

Pro's:

-easy construction

-high resolution scanning possible (I think)

- cheap

-....

Con's:

- limited volume that can be scanned

-....

I would like to see your comments

[Titus 170]

Love the idea! Some parts could even be printed if accuracy allows for it. I'm in for working on this one!

@cons: volume is limited, but if you pick the parts so that it could also be mounted outside of the frame(i.e. design/sell a bigger housing) than you have both the easy and cheap, and the larger stuff!

 

[nallath 1,123]

Depends on your definition of high resolution. With a laser in most situations you will get an accuracy of about 0.5 to 1 mm, unless you start using high end (industrial) lasers & camera's.

Cons:

 

• Every time you move the scanner from your printer, you will have to re-do the calibration process. This can take about 5 minutes.
  
• With no moving laser, there are large parts that are 'un-scanable'
  
• The printer is white. Which is the worst color for 3D scanning background ever.
  
• It's not that cheap. You will need a decent camera (100-150 euro), a laser bandpass filter (80 euro), a semi decent laser (20 ish euro) and a rotation platform (20-50 euro).
  
• Distance of the camera to the to-be scanned object is small, which forces you to use cameras with a large angle, which tend to have a lot of fish eye effect. This is killing for your quality.
  

 

[peggyb 171]

I don't understand why you want to put it inside the printer...

The turntable can stand everywhere (no size limit there), the calibration panel is just 2 panels at a 90 degr. angle, so they will stand by themselves and are easy to store. The camera can be on a tripod..

You will have to buy the software anyway.

 

[Titus 170]

Perhaps just storable in the printer?

 

[th3o 0]

The idea is if you put it inside the printer 1)you can make use of the bed, so you can let it go up 1 step at the time, so you can make accurate scans

2) you can put a black cloth over it so you have a dark box to scan in.

3) by using the case of the Ultimaker you can make the bracket and turntable removable but mount it in a way the parts are always at the same position. Making it easier to calibrate.

But maybe I'm wrong. The whole scanning stuff is new to me.

About the background: you can make a black cardboard or plastic background

[nallath 1,123]

All true, but all of this would be much, much easier (and better to set-up) if you simply made another box instead of relying on the UM encasing.

 

[peggyb 171]

see here:

the whole setup in a box

 

[Daid 306]

The idea is if you put it inside the printer 1)you can make use of the bed, so you can let it go up 1 step at the time, so you can make accurate scans

2) you can put a black cloth over it so you have a dark box to scan in.

3) by using the case of the Ultimaker you can make the bracket and turntable removable but mount it in a way the parts are always at the same position. Making it easier to calibrate.

But maybe I'm wrong. The whole scanning stuff is new to me.

About the background: you can make a black cardboard or plastic background

 

So... you mount it in a box? Little reason why it needs to be an Ultimaker. Ikea has lots of cheap options that would fit better. The UM casing is also a bit small for a scanner.

(The bed has little added value, as moving up/down moves you out of the focus or visible range of the camera)

 

[nallath 1,123]

Not quite true. It would have some added value. If you place the camera outside the printer (perpendicular to the ground plane) it would help. You could stack multiple lower res scans together like that.

 

[th3o 0]

@nallath That is what was the idea. At this moment I have a camera that I use for timelapse. It is mounted at a stationary position and I don't have any trouble with focus.

 

[peggyb 171]

you want to use a laser (not a beamer?)

That way you could (?) use the z stage to move the object up and down instead of moving the laser...

 

[th3o 0]

Yes. That is the idea. So you can have a fixed bracket for the laser and the camera

 

[peggyb 171]

If you do it that way, it is important that your model doesn't vibrate during the movement of the z-stage.

 

[nallath 1,123]

And it won't scan all area's of the model visible to the camera. The results would be far better with a projector (or rotating laser, but that creates a whole range of extra problems)

 

[conz 0]

Maybe a 3d-scanner inside the printer could be used (somwhere in the future) to monitor the printing.

It could detect failed prints where you get rat-nests, shifting, clogged nozzle and similar problems to stop the print.

With really high resolution you could detect underextrusion, too?

 

[nallath 1,123]

Yes, but it's not a trivial thing to do. I'd put my money with 'ordinary' computer vision algorithms if you want that sort of stuff.

All 3d scanning algorithms have several tradeoffs; Complexity, robustness, number of frames required, processing time, accuracy and resolution.

In order to get scans that are good enough to detect under extrusion, you would need a pretty awesome resolution and accuracy. It would also need to be extremely robust, as false positives / negatives are a big no-no. This usually means that you require either a ton of frames and/or a ton of processing time and or huge complexity. But then again, even a ton of frames is kind of a no-no, and the same can be said for a ton of processing time, leaving only one thing to be unoptimized; The complexity.

 

[th3o 0]

And it won't scan all area's of the model visible to the camera. The results would be far better with a projector (or rotating laser, but that creates a whole range of extra problems)

 

With the use of the turntable I think the only side that is not scanned is the underside. But I think that will always be a problem...

 

[nallath 1,123]

Not quite. The same problem occurs with the makerbot digitizer. Only if you sweep the laser (and rotate the platform) will you get the largest scanned area.

 

[th3o 0]

Why do you have to sweep the laser if you move the Z-stage?

 

[nallath 1,123]

I cannot explain this well with only using words, but basicly your problem is self occlusion.

 

[th3o 0]

OK. I think I understand what you mean

 

[th3o 0]

There is someone who had the same idea; http://www.thingiverse.com/thing:253770/#comments

In this case he isn't using a turntable, so it only the frontside is scanned.

 

[nallath 1,123]

Interesting, although it doesn't really add anything to the printer. I'm greatly in favour of the 'if you stick two thing together, it should improve both camp'. I think this only saves you a camera mount.