Our Gift to You: 20% off your own designs in Strong & Flexible Plastic through Dec 23, with promo code MY2014 · details
Home » Support » 3D Printing » laserscan vs photogrammetry
Search Search  
Show: Today's Messages    Show Polls    Message Navigator
laserscan vs photogrammetry [message #85213] Fri, 21 February 2014 16:57 UTC Go to next message
avatar jackojag  is currently offline jackojag
Messages: 2
Registered: February 2014
Go to all my models
Junior Member
Hello,

Can someone help me decide which way to scan people?
I would like to make figurines but i want them to be clear and colorfull.
Is there a lot off difference( in color and clearness ) in using like : Artec Eva, Mephisto ex-pro,creaform goscan
or make a scan making use off photogrammetry and multiple dslr-cameras like : XXarray ( 62 cameras ).
Thanks in advance
Re: laserscan vs photogrammetry [message #85328 is a reply to message #85213 ] Mon, 24 February 2014 02:05 UTC Go to previous messageGo to next message
avatar ThreeForm  is currently offline ThreeForm
Messages: 22
Registered: March 2012
Go to my shop
Junior Member
In general, photogrammetry will have sharper color detail, and the other methods will have better geometric detail. Photogrammetry can use one or many cameras, and the DSLR array will have the advantage of rapid exposure, so that the movement of the subject doesnt reduce quality. The other methods you mention are variations of structured-light scanning (though creaform does use lasers to generate the pattern). I consider them medium detail because accuracy is limited by sensor pixel size. Some versions include color. Their hand-held format makes them a good fit for body scanning, but the cost is far from hobby-level. Mephisto is higher quality type of scanning (phase-based structured light, a.k.a. white-light scanning) with a longer single exposure but great color and detail. Phase-based scanning can capture a sample for every pixel in the image sensor, which quickly produces millions of points. They have a depth accuracy up to about 1/10th the pixel size (10x more accurate in Z than in X and Y). Like DSLRs, structured light scanners can be arrayed to get rapid full scans, but this is a six-figure cost and not portable.

There are several hobby-level methods the use the PrimeSense sensor, but since you are looking for quality, do not waste your time on any of that. They are a tenth of the cost, but a tenth of the quality.

Structured light would be my recommendation because of its flexibility, but expect to pay at least $25,000, probably $35,000+ for the hi-res color systems. Otherwise if you are in a studio and not on site, photogrammetry is probably the best choice because you can start with just a few cameras and scale the system over time. Agisoft Photoscan is one of the software packages to do this.

[Updated on: Mon, 24 February 2014 07:09 UTC]


www.facebook.com/pages/ThreeForm/114001098701667
Re: laserscan vs photogrammetry [message #85358 is a reply to message #85328 ] Mon, 24 February 2014 13:35 UTC Go to previous messageGo to next message
avatar jackojag  is currently offline jackojag
Messages: 2
Registered: February 2014
Go to all my models
Junior Member
Thanks for your reaction,
It is a difficult decision, because i want my scan to be good, and i would like to scan as fast as possible ( regarding my budget 25.000 euro ).
Now i am thinking to make an array of 10-15 dslr cameras ,work with a turntable, take about 5 shots, and ready.
I think this will go faster, more accurate and cheaper then by doing the scan with the Mephisto Ex-pro.
Am i right to think this?
Gr. Jack
Re: laserscan vs photogrammetry [message #85397 is a reply to message #85358 ] Mon, 24 February 2014 23:27 UTC Go to previous message
avatar ThreeForm  is currently offline ThreeForm
Messages: 22
Registered: March 2012
Go to my shop
Junior Member
I think you're on the right track. While the Mephisto might be sharper, the person will move so much between shots that it won't matter. With a rapid series of flash shots with your limited DSLR array, you can at least discard the first or last shots if there was movement.

You may actually be able to get an all-around shot with a limited array. There are rapidly diminishing returns as you add cameras. They are mostly needed to fill in occlusions between the thighs and under the arms. Each doubling of the number of cameras will increase the completeness by a few percent, and you will never get everything. Rather than spending it all on cameras, get some good software for cleanup. Zbrush is pretty popular since it is reasonably priced and has good tools for rebuilding textures.

The cameras also don't necessarily have to be high-end Canons or Nikons or even real SLRs. They just need good manual control, remote triggering and high enough build quality to ensure the optics are consistent between units, since the algorithm assumes they are completely identical.


www.facebook.com/pages/ThreeForm/114001098701667

 
   
Previous Topic:Printing with spures (intakes) for raw brass
Next Topic:Are deviations repeatable?

Logo

Hello.

We're sorry to inform you that we no longer support this browser and can't confirm that everything will work as expected. For the best Shapeways experience, please use one of the following browsers:

Click anywhere outside this window to continue.