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

Aspest Ratio and Scan Acuracy and Appature Distortion

Problems, solutions and discussion about scanning with a video projector and stripe patterns.

Aspest Ratio and Scan Acuracy and Appature Distortion

Postby micr0 » Mon Feb 01, 2016 4:59 pm

I've recently been on a mission to improve my scans. The release of the SLS3 kit and CuriousJeff's experiments with 4K confirmed my suspicions that higher resolution hardware was going to be an easy way to get significant improvements. I was able get for a one month evaluation imagingsources new DMK 23UX174 Camera, which is a USB3 monochrome camera with a resolution of1920x1200, and an LG pf1500 projector which is a Full HD (1920x1080) LED projector. The LG was also my choice because it is considered portable, so it's small, but most importantly it doesn't have diamond pixels but instead used square patterns which in and of it's self give better performance with David. I don't have the camera yet, but I have had the projector for 2 days now and wow what a difference. For the record I had been using an Optoma ML550 which was far better than the Acer K330 I started out with. I'm using a David Cam3M. Just the addition of this projector has given me much improved results with the horizontal lines (vertical if your camera set to the side of the projector) almost completely now gone. For all intents and purposes they are gone and I only saw them very faintly when I scanned a flat white glass panel as a test. I started testing this new set up by re scanning a cylinder head I had already scanned with my old projector. The detail was noticeably clearer with small casting marks and features much more clearly defined. However now that I was getting better scans I started to pay closer attention do a few things that I have previously not thought much about. The pictures Tibor posted a day or so ago, where I remarked that it looked like he had an alignment issue, also came to mind and got me thinking. I use ping pong balls as alignment references and was noticing that no matter carefully I aligned the scans The balls in the scan never seemed to be perfectly round. In fact some of the ball surfaces/scans didn't line up very well at all. This new projector is brighter than the Optoma and I was able to stop the aperture of the camera further down. I though this was good as I was increasing the depth of field and therefore the depth of the work envelop should be larger. However now I'm suspecting two things. first is that even though the camera has a greater depth of field, the projector does not. that is that the focus of the projected image must very depending on the distance of the feature being scanned. However I am suspecting that the real culprit of my problems is actually Diffraction caused by the small aperture opening. I'm going to test this theory out today and see if opening the lens up helps or not.

My questions are:

Has anyone seen distortion caused by diffraction in the lens at small aperture settings?

Does anyone know what the formula for figuring the work envelope is? I know it is related to camera and projector distance from the subject but what is that ratio? At this point I'm most interested in the depth dimension for features at are closer to and further from the camera.

Thanks


µ


This is a quick scan with the new projector. I'll try to find an image of the previous scan of this head for a before comparison.

Smart Head 1-31-2016.jpg
µ
micr0
 
Posts: 315
Joined: Wed Mar 19, 2014 3:54 pm
Location: NYC

Re: Aspect Ratio and Scan Acuracy and Apeture Distortion

Postby micr0 » Mon Feb 01, 2016 5:14 pm

I also wanted to ask:

I'm assuming that the projected patterns were made to be projected from a projector with a 16:10 (K132 1280X800) aspect ratio. A projector projecting at a 16:9 aspect would cause a slight stretching of these patterns. How much does this effect the accuracy of the scans?


µ
µ
micr0
 
Posts: 315
Joined: Wed Mar 19, 2014 3:54 pm
Location: NYC

Re: Aspest Ratio and Scan Acuracy and Appature Distortion

Postby Curiousjeff » Tue Feb 02, 2016 7:55 am

Great scan ! I love the surface.

I don't know the answer regarding various distortions (lens quality, f-stop diffraction, chromatic abberation, etc...) and the technique of DLS to compensate but a good starting point would be to scan the calibration panel just after calibration and make sure there is no or little distortion at this early stage

http://forum.david-3d.com/viewtopic.php?f=1&t=6194#p30328
http://forum.david-3d.com/viewtopic.php?f=1&t=398&p=1808&hilit=Check+corner#p1808

I am interested in your settings. For the image you posted, how many patterns were projected ?

Jeff
Panasonic GH4 4K with lens 12-35, Inogeni 4K to USB3 hdmi capture device, projectors: Acer K135/Acer K11/NEC V260X, Canon 60D, macro 60 mm, MagicLantern
Curiousjeff
 
Posts: 230
Joined: Fri Feb 27, 2015 2:17 pm
Location: Geneva, Swizerland

Re: Aspest Ratio and Scan Acuracy and Appature Distortion

Postby Jokeri » Tue Feb 02, 2016 12:13 pm

Micr0,

could you please scan your calibration panels and post the file here?

I would like to compare it to my panel scans.
Jokeri
 
Posts: 36
Joined: Sun Mar 27, 2011 3:24 pm

Re: Aspest Ratio and Scan Acuracy and Appature Distortion

Postby micr0 » Tue Feb 02, 2016 3:50 pm

Those scans were 22 patterns. I have scanned my calibration panels and they are good (flat and square) to the eye. I'll post a pic later today for reference. The problem I was having is that the scan at what would be the outer limits of the scan envelop in the dept dimension are not lining up during alignment. This deviation is very slight and I used to just chalk it up to acceptable error in the scans. However I started thinking about it as I have had scans where it didn't exist. It got me thinking about how that error projects through the mesh, or more importantly how accurate and true the entire mesh is to the original part, but I haven't tried to measure it yet.
That said, yesterday I did have time to do some scanning and the first thing I tried was reducing the projector brightness and opening the aperture of the camera and it worked. I can say with confidence that the problem I was experiencing was due (at least mainly) to diffraction caused by a small aperture. The scans I did yesterday (the exact same set up as before) now align almost perfectly.

The spheres in the scans are ping pong balls I use for alignment purposes. You can see in the image below that on the right rear sphere, one of the surfaces doesn't quite seem to line up with the others. This is the indicator of the problem. That deviation is slight and previously I'd get it as close as possible and during fusion it would look fine. However the engineer in me knows it's off. I think i have a screen shot of the scans I did after I opened the aperture and I'll post then as soon as I can edit them for size.

µ

Align 1.jpg
µ
micr0
 
Posts: 315
Joined: Wed Mar 19, 2014 3:54 pm
Location: NYC

Re: Aspest Ratio and Scan Acuracy and Appature Distortion

Postby Curiousjeff » Tue Feb 02, 2016 6:09 pm

Since you are "only" using 22 patterns, I wonder if the improvement is due to the higher resolution of the projector or only to the non diamond alignments of the pixels?

Jeff
Panasonic GH4 4K with lens 12-35, Inogeni 4K to USB3 hdmi capture device, projectors: Acer K135/Acer K11/NEC V260X, Canon 60D, macro 60 mm, MagicLantern
Curiousjeff
 
Posts: 230
Joined: Fri Feb 27, 2015 2:17 pm
Location: Geneva, Swizerland

Re: Aspest Ratio and Scan Acuracy and Appature Distortion

Postby edegre » Wed Feb 03, 2016 12:22 pm

Hi everyone,

I'm posting for the first time, but I follow your posts since a long time, and I'm quite impressed by your "researching" and in fact that's one of the reasons why I have deciced to chose DAVID's solution. I did it some years ago for my former company, and I do it again for the company I've just created. So THANKYOU guys !!

I'd like to check with you if some of you have tested the camera DMK23UX174, 1900x1200, 54fps, USB3, Directshow. It seems to have similar caracteristics to DAVID-CAM-4, but obviously withou the propietary firmware. Does some of you know what would this change in practice?

It's expensive for my bugdet, but the promise of doubling the resolution and the speed of scan is very tempting...

Thank you in advance for your replies, and I hope to be very soon in a position to share my own scans.

Eduardo de Gregorio
Sculpture-Lab /Paris
edegre
 
Posts: 5
Joined: Tue Oct 20, 2015 7:25 pm


Return to Structured Light Scanning

Who is online

Users browsing this forum: No registered users and 1 guest

cron