Projecting an SLS system: Scanning precision

Write here about construction, ideas, equipment, tips n tricks etc. related to structured light scanning
User avatar
Micr0
Posts: 586
Joined: 15 Nov 2016, 15:20
Location: New York City

Re: Projecting an SLS system: Scanning precision

Post by Micr0 »

3dxcan wrote: 17 Nov 2017, 13:46
mading wrote: 17 Nov 2017, 09:24 Hi to all, I gave a tried to GOM inspect 2016, free version:

I have to check the perpendicularity of my panel: it is measured as 90.05°:
90.05°.jpg

Then I can fit planes (easy and straighforward):
Clip_4.jpg

Here you have deviations from fitted planes:
Clip_2.jpg Clip.jpg
as you see in the pics, you have a center of yellow and green (~0.05 divation) which happens to be accelerating out of the divation zone (-0.15) as you move toward the calibration edges, gray areas on the corner are our of tolerance zone (-0.15,0.15).Its a lens problem. someone with a decent lens should give us his result.
what size of calibration panel are you using? whats the format of the camera sensor (eg 2/3", etc) and lens? what type of camera? global/rolling?
Also understand that even .15mm deviation is still within the volumetric accuracy of systems costing $20K+ I have an Artec Eva here on loan and I get .15mm on a good day. What I find is that the real difference is that the Artec software handles this much better than David. Artec has better error compensation algorithms so that you generally don't have fusion problems. That said I get clean smooth models that are easily .15mm out when compared to hard measured machined surfaces. Artech studio is very good a weighting the accuracy of the raw scanned data and how it influences the fused model. I went down the rabbit hole of trying to get David perfect and got very close. But I spent HOURS systematically doing it and I cant say after using David practically for a few years that it was completely worth it. I have also not found that level of precision to be easily repeatable over different scale setups. What I'm saying is don't spend too much time chasing perfection till you actually get some use out of the system and find out what your actual needs are. The more you use it the better you will become at understanding the errors in your system.
µ
mading
Posts: 307
Joined: 31 Jan 2017, 13:09

Re: Projecting an SLS system: Scanning precision

Post by mading »

Many thanks Micro.
I was thinking to ask you about the 2nd Cam.
Although I guess you already had very good results aligning scans with a single cam, do you think a newby will benefit for having a 2nd cam in extended mode? Have you found easier to align with two cam, especially a complex free-form 360° shape?
Will it make easier to avoid alignement mistakes?
LG PF50, LG PF1500, RangeVision DIY: 2x DahengMer630, 2X12 and 2X16 mm 5Mp ZK lenses, RV turntable
mading
Posts: 307
Joined: 31 Jan 2017, 13:09

Re: Projecting an SLS system: Scanning precision

Post by mading »

Dear All, I decided to switch to a simle form for my testing. I realized alignement needs improvements, so I practiced on a piece of wood.
2017-11-21_09-17-24.png
Dimension are in good agreement with my caliper up to 0.08-0.14 mm.
It's not easy to take measurement on stl :-)
So I think It's a starting point.
2017-11-21_09-17-21.png
Here you have 6 scans, calibration panels 49 mm, the block is around 43 mm.
-Scan 0, 0.2, 0
-Fusion 1500, 2, 1%
2017-11-21_09-25-45.png
LG PF50, LG PF1500, RangeVision DIY: 2x DahengMer630, 2X12 and 2X16 mm 5Mp ZK lenses, RV turntable
mading
Posts: 307
Joined: 31 Jan 2017, 13:09

Re: Projecting an SLS system: Scanning precision

Post by mading »

Here we go: I start understansing a little bit more how to treat some little imperfections.
Attachments
2017-11-21_16-27-29.png
2017-11-21_16-27-16.png
LG PF50, LG PF1500, RangeVision DIY: 2x DahengMer630, 2X12 and 2X16 mm 5Mp ZK lenses, RV turntable
Post Reply