Dear DAVID Community,

As you are probably aware, the entire DAVID company, team and product range has been integrated into HP Inc.
This forum is now read-only. Relevant content has been migrated to HP and merged into HP's Support Forums on November 1, 2016.

To start new discussions for 3D Scanning, please register and post your new topic at the HP Support Forums

[Gallery] SL results

Discussions about latest DAVID developments, beta versions...

Postby hal » Mon Sep 19, 2011 10:29 pm

Hello and thanks.

@ Sam: what can I say is that, for my experience, with the same camera, Laser and SL achieve similar quality of scans. I don't see big difference of details or other (if working with same workflow). But the "boost" of the SL technique is, obviously, the speed. :mrgreen:
So, you can achieve similar/equal results if work with laser. Maybe is better to coat the object, with this tecnique. SL is stronger than the Laser survey when it must to read correctly the projected pattern on the object. So with SL isn't so necessary to coat (take in mind that with shiny and very dark surfaces, if you want perfect scans, coating is also a must).

I took around 30 single scans (6 seconds each one, so we are talking around 3 minutes of pure scanning!) for the sandal (because the upper and lower shoelaces have many undercuts and hidden areas).

The original fused mesh (with David 3 at value 1024) count 3.500.000 triangles. The dicemated version that you see in the renders count around 600.000 triangles.

Well, as ever in my workflow, post-production is a big part of the whole job. In this case I've divided the object in 3 pieces: Sole, Upper Shoelace and Lover Shoelace. Than I've aligned the scans that cover the "shadow areas" and performed the Global Fine Registration for every pieces. Than the 3 object were fused separately at 1024 of value.

I think that scanning, saving, loading, re-saving, aligning different time, again saving and finally fusing... we are talking around 40/60 minutes of work (and I do this in my spare time and usually in the night, after daily work, so, my mental reflex and concentration isn't at 100% :roll: )

Hope that these are good answers for you.
Regards
Mattia
User avatar
hal
Moderator
 
Posts: 1527
Joined: Thu Mar 01, 2007 4:08 pm
Location: Cavaria con Premezzo (VA), Italy.

Re: [Gallery] SL results

Postby icstra » Fri Sep 23, 2011 4:12 pm

Hi Gunter,
If I've well understood your post viewtopic.php?f=16&t=3060&start=30#p17463
when you divide the vertical resolution of the projector by the number of sine waves ( depending on frequency), you obtain the maximum number of shifts you can get.
So, now, my projector has a vertical resolution of 768.
It means that at a frequency of 8, the maximum theroical number of shifts is 12. ( 768/64)
My question is the following :
what are the best settings to follow the theorical rule seen above ?
Set frequency at maximum and calculate max shift or set max shit and calculate frequency?
I don't know if my question is really clear but hope however a clear reply :-)

Thierry
MacBookPro 17" I5 Dual Core 2xSSD 128Go, RAM 8Go, OSX Lion, Bootcamp, Win7 X64
EOS 550d and Logitech HD Pro C910
LG HX300G 1024x768

www.icstra,ch
icstra
 
Posts: 44
Joined: Sat Sep 10, 2011 9:12 am
Location: Châtel-St-Denis Switzerland

Hal's Response

Postby shoesam » Fri Sep 23, 2011 8:44 pm

Mattia,

thank you so much for answering my questions.....the hope you have given me and the insight as to how I can acheive these sort of results is much appreciated.

Sam
shoesam
 
Posts: 39
Joined: Tue Jun 07, 2011 7:36 pm

Re: [Gallery] SL results

Postby MagWeb » Fri Sep 23, 2011 10:15 pm

icstra wrote:Hi Gunter,
If I've well understood your post viewtopic.php?f=16&t=3060&start=30#p17463
when you divide the vertical resolution of the projector by the number of sine waves ( depending on frequency), you obtain the maximum number of shifts you can get.
So, now, my projector has a vertical resolution of 768.
It means that at a frequency of 8, the maximum theroical number of shifts is 12. ( 768/64)
My question is the following :
what are the best settings to follow the theorical rule seen above ?
Set frequency at maximum and calculate max shift or set max shit and calculate frequency?
I don't know if my question is really clear but hope however a clear reply :-)

Thierry


With the post you linked to I wanted to point out, that you can get equal waves, only if your frequency fits to your resolution. This is theoretical. In praxis you will find that visible in only a few (maybe never) situations, for the projection will always be a bit unsharp in depth of 3D. ( Nevertheless I think that it doesn't make much sense if you set a high frequency if you own only a poor projector)
I have the same vertical resolution. If the scanning time does not play a role, I use frequency 7 or 8 (depends on the surface) @ 16 phaseshifts. Your maths are right (12 shifts), but this way I got the best results.... But I´ll try 12 shifts at frequ 8 next time.

Gunter
MagWeb
Moderator
 
Posts: 2499
Joined: Wed Jul 18, 2007 8:48 pm

Re: [Gallery] SL results

Postby WalterMo » Fri Oct 07, 2011 3:31 pm

I am just testing the resolution for close by object scans with my (only) 800 x 600 pixels projector. And why not to use something which was already scanned before. I found the model of an Opel veteran car from 1909 in my "fundus". 80mm long.
And here is the scan, made exactly two years ago:
viewtopic.php?f=2&t=1295&p=8715&hilit

Used hardware this time: DAVID monochrome CCD camera, Acer K11 projector. The projector was equipped with an additional convex lens (f=400mm) to enable focussing from 26 cm to 52 cm. Outer calibration corner, scale length: 36.2mm.

Settings:
Coded light: inverse mode
Phase Shifts: 12
Frequencies: 8
Orientation: both
Pattern holding time: 300ms
No interpolation or smoothing

I remember, at that time I needed about 3 minutes for a single scan and now only 15 sec. :D

Walter

Edit: Only the third photo shows a fusion of 17 single scans (resolution = 600). Both first are single scans each.
Attachments
Opel Scan Screen dump.jpg
Opel Scan Screen dump v unten.jpg
Opel Screen duo.jpg
Scanned on a turntable. Rotated on x- and y-axis. Combined of 17 single scans.
User avatar
WalterMo
Moderator
 
Posts: 2363
Joined: Mon Apr 02, 2007 6:52 pm
Location: Braunschweig, Germany

Re: [Gallery] SL results

Postby WalterMo » Sat Oct 08, 2011 4:23 pm

If the setup is once built-up it's easy to continue with another object. This time I have chosen a broken bulb from a car's front lamp.

Same hardware and same settings as above. Again I used a turntable with 8 positions á 45°.

Walter

PS. DAVID's SL is really great. That's fun! :D
Attachments
Autobirne Screendump.jpg
Fused with a resolution of 600
Aufbau SL Birnenscan.jpg
Used setup
User avatar
WalterMo
Moderator
 
Posts: 2363
Joined: Mon Apr 02, 2007 6:52 pm
Location: Braunschweig, Germany

Re: [Gallery] SL results

Postby Khalid Khattak » Sun Oct 09, 2011 4:54 am

Walter nice rig but the scan result is not that clear.. What Projector are you using.. Can you also tell me what projector Mattia using?.. I am in search of purchasing a suitable projector...
Edit:
I am watching a Controller box at the Left of the Rig with green backlit LCD.. Is this combine (Turntable+SL controller) or what?
PC: Dell Inspiron i3, 2.27GHz,4GB RAM,64 bit windows 7
CAMERA:
David CCD Mono USB
LASER:
David 5mW Green line laser
PROJECTOR:
3D LED Mini Projector 250lumens, Native 1024x600
[b]http://free3dscans.blogspot.com/
Khalid Khattak
 
Posts: 942
Joined: Thu Apr 23, 2009 1:30 pm
Location: Pakistan

Re: [Gallery] SL results

Postby WalterMo » Sun Oct 09, 2011 8:11 am

Hi Khalid,
I think the single scans are clear. See the attached scan. The filament of the bulb was spayed with a light grey colour some years ago and the coil was therefore smeared. Nowadays I would use this spay (May 11th 2010):
viewtopic.php?f=2&t=1283&start=15

And the filament is positioned in the focus of the bulb's reflector, means the projector's light is also focused from behind to the coil, even if the reflector was spayed.

Also, I am not a specialist in stitching of several scans of such small parts. :oops:

I am using an ACER K11 projector. Here you can see what other guys are using for SL scanning:
viewtopic.php?f=16&t=3044
Mattia hasn't mentioned his projector there, but it is an Adapt ADPP-305.

My controller for the turntable is a stand-alone device.
Attachments
Einzelscan Birne.jpg
Single Scan
User avatar
WalterMo
Moderator
 
Posts: 2363
Joined: Mon Apr 02, 2007 6:52 pm
Location: Braunschweig, Germany

Re: [Gallery] SL results

Postby Byung-gyu » Tue Oct 11, 2011 1:45 am

MagWeb wrote:Hi Mattia,

you´ve to do something against reflections inside the corner(from one plane to the other). Either you use a external corner, as I do, or you put a thin black plane in the direction of the projector (45° to the corner) so that one side of the corner cann´t "see" (or reflect to) the other...

Make sure that your machine is set to 96dpi (100% symbol size). Don´t use large, 120% symbol size...

Gunter


Hello Gunter,

Your posts here are a great help to many people including me. thank you again.
I have read over and over again to understand and to get me to be like you :D
here is question, what you mean by "set to 96dpi(100% symbol size)"?
If i'm right this means that projector's zoom funtion shouldn't be used?
Could you give me more detail?

Thanks in advance
User avatar
Byung-gyu
 
Posts: 24
Joined: Sun May 31, 2009 3:29 pm
Location: Seoul, South Korea

Re: [Gallery] SL results

Postby WalterMo » Wed Oct 12, 2011 8:33 pm

I am quite sure that Gunter meant the settings of the second monitor (video projector) of the graphics card. The normal setting is 96 dpi (dots per inch). It is possible to increase this value to 120 dpi if elements, e.g. symbols) are shown too small on the PC screen.

Walter
User avatar
WalterMo
Moderator
 
Posts: 2363
Joined: Mon Apr 02, 2007 6:52 pm
Location: Braunschweig, Germany

Re: [Gallery] SL results

Postby MagWeb » Wed Oct 12, 2011 11:36 pm

Hi,

Walter is right :D

But this is an issue of the past. Nowadays DAVID detects your used dpi settings correctly . You don´t have to worry about anymore. (Sometimes Beta testers can contribute to ease your life :wink: )

You may use the optical zoom of your projector - no problem-, if you do not change it after projector calibration. Remember: Narrow angles of projections (or the cam's field of view) mean less distortions of projection ( or the image) . And this means less work for DAVID to correct them (which is allways a source of error).

Gunter
MagWeb
Moderator
 
Posts: 2499
Joined: Wed Jul 18, 2007 8:48 pm

Re: [Gallery] SL results

Postby WalterMo » Thu Oct 13, 2011 8:00 pm

And here is my first SL scan of a coin. It's the same coin which I had used to test a new spray to tarnish the silver surface:
viewtopic.php?f=2&t=1283&start=15

As opposed to the scans of the model veteran car and the bulb above I have used my modified Pro9000 with a cheap zoom lens. The resolution was set to 1280 x 1024 pixels. Used DAVID: V 3.1.
The video projector (Acer K11) was equipped with a 130mm lens in front of its lens to enable nearby focussing. Outer calibration corner with 36.2mm scale length.

Settings:
Coded light: inverse mode
Projector brightness was reduced (by the new slider) to only 40%.
Phase Shifts: 12
Frequencies: 8
Orientation: both
Pattern holding time: 500ms
Not interpolated and not smoothed
Total number of patterns: 52

Walter
Attachments
DuoBildMünze3.jpg
Front- and backside of a 5 DM silver commemorative coin. 29mm in diameter.
User avatar
WalterMo
Moderator
 
Posts: 2363
Joined: Mon Apr 02, 2007 6:52 pm
Location: Braunschweig, Germany

Re: [Gallery] SL results

Postby WalterMo » Mon Jan 23, 2012 6:53 pm

My good old Nokia 3410, scanned on a turntable (8 x 45°).

Camera: Logitech C601 at 800x600 with a zoom lens of 6-15mm:
viewtopic.php?p=20118#p20118

Projector: Acer K11 with a lens of f = 400 mm in front. In this way I got a focus range of 26-52cm.
Inner calibration corner. Scale: 145mm.
Fusion: 800.

SL advanced settings:
Before you change frequencies and pattern holding time you have to switch its automatics off! Otherwise DAVID will hang. :(
Pattern Holding Time: 300ms.
Frequencies: 8
Phase Shifts: 12

Walter
Attachments
Nokia Panoramabild.jpg
User avatar
WalterMo
Moderator
 
Posts: 2363
Joined: Mon Apr 02, 2007 6:52 pm
Location: Braunschweig, Germany

Re: [Gallery] SL results

Postby WalterMo » Tue Mar 27, 2012 5:41 pm

And now the Nokia as a Flash presentation.
The original *.obj file of 46MB was reduced by the reduction tool of Shape Fusion (set to 6%) to 1.83MB. The appertaining *.png file for the texture was let as it was at 1.3MB.




User avatar
WalterMo
Moderator
 
Posts: 2363
Joined: Mon Apr 02, 2007 6:52 pm
Location: Braunschweig, Germany

Postby hal » Tue Mar 27, 2012 8:54 pm

Very nice :D
Thanks for sharing.

Mattia
User avatar
hal
Moderator
 
Posts: 1527
Joined: Thu Mar 01, 2007 4:08 pm
Location: Cavaria con Premezzo (VA), Italy.

PreviousNext

Return to DAVID Development

Who is online

Users browsing this forum: No registered users and 1 guest