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

Structured light scanning failing??

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

Structured light scanning failing??

Postby karlkohl » Mon Dec 13, 2010 12:57 pm

I have done the setup as per mention in the forum and the David scanner manual.
Performed the calibration steps and took proper images.
Yet structured light scanning is not working.

What could be going wrong??
Any clues what can I do right to get a decent scan??
The calibration result was successful but even after that I can't get a proper scan.

My work-flow is as under.
Projector : 1024 x 768 DLP
Cam: 14Mp DSLR with 18-55mm lens ---> outputs RAW images (4608 x 3072pixels)

I am resizing the images to 1152 x 768px and converting to ".bmp" format to maintain the aspect ratio(3:2) of the original image and also so that David can work with them.
Is this causing the problem.

I have upload all my images (File name 1.rar), low res jpg's.
Can't upload the originals as the forum doesn't allow me to upload large attachments.
Attachments
slimgs.zip
(1.42 MiB) Downloaded 182 times
karlkohl
 
Posts: 26
Joined: Sun Dec 12, 2010 8:14 pm

Re: Structured light scanning failing??

Postby MagWeb » Mon Dec 13, 2010 1:22 pm

Hi,

at a first quick look:

- move your cam closer to the corner, so it sees nothing else but the corner;
- fix your patterns absolutly flat on the corner (best: use selfadhesive paper);

EDIT:
- increase contrast ,so the images are nearly black in the dark zones of the projections.
EDIT:
- do not use a totally back image instead of the image of the totally black projection (img01)
MagWeb
Moderator
 
Posts: 2499
Joined: Wed Jul 18, 2007 8:48 pm

Re: Structured light scanning failing??

Postby karlkohl » Mon Dec 13, 2010 2:05 pm

Hi,

at a first quick look:

- move your cam closer to the corner, so it sees nothing else but the corner;
- fix your patterns absolutly flat on the corner (best: use selfadhesive paper);

EDIT:
- increase contrast ,so the images are nearly black in the dark zones of the projections.
EDIT:
- do not use a totally back image instead of the image of the totally black projection (img01)


I have to increase the contrast of the projector right?
And I'm not quite sure what you meant by saying "do not use a totally back image instead of the image of the totally black projection (img01)"?
karlkohl
 
Posts: 26
Joined: Sun Dec 12, 2010 8:14 pm

Re: Structured light scanning failing??

Postby MagWeb » Mon Dec 13, 2010 2:54 pm

It doesn´t matter if whether you increase the projector's or cam´s contrast - important is the resulting image, which should be as dark as possible where no light was projected onto the scene.

Your img01 is totally black all over - seems there are no gray values at all. In the past, some lazy guys simply replaced that image by a virtually made black one. It is important that you actually shoot this image with the same settings of the cam as used in all other images. This image defines the "totally black status" and its values are "subtracted" from the following images... It seems to me that your img01 does not fit to all the other images.
MagWeb
Moderator
 
Posts: 2499
Joined: Wed Jul 18, 2007 8:48 pm

Re: Structured light scanning failing??

Postby karlkohl » Mon Dec 13, 2010 3:09 pm

MagWeb wrote:It doesn´t matter if whether you increase the projector's or cam´s contrast - important is the resulting image, which should be as dark as possible where no light was projected onto the scene.

Your img01 is totally black all over - seems there are no gray values at all. In the past, some lazy guys simply replaced that image by a virtually made black one. It is important that you actually shoot this image with the same settings of the cam as used in all othher images. This image defines the "totally black status" and its values are "subtracted" from the following images... It seems to me that your img01 does not fit to all the other images.


I had kept the same setting's of the cam for all the images also the cam was set to manual mode.
I think while the images have somewhat changed due to reduction in quality for uploading to the web, not sure any other reason for it to look that bad.
Do you think my badly made calibration corner is the root of all problems?
However calibration was successful in the software the crosshairs of the left image were off target.
Attachments
calresult.jpg
karlkohl
 
Posts: 26
Joined: Sun Dec 12, 2010 8:14 pm

Re: Structured light scanning failing??

Postby MagWeb » Mon Dec 13, 2010 3:53 pm

Yes, besides the wavy fixing of the patterns ( I mentioned above ) , it seems that the distance of the corner to the inner rows of points is a bit smaller.

Your boards seem to be corrugated cardboard ? Choose some realy flat material (white furniture board or something similar) and set the 90°angle carefully.
MagWeb
Moderator
 
Posts: 2499
Joined: Wed Jul 18, 2007 8:48 pm

Re: Structured light scanning failing??

Postby karlkohl » Mon Dec 13, 2010 5:54 pm

MagWeb wrote:Yes, besides the wavy fixing of the patterns ( I mentioned above ) , it seems that the distance of the corner to the inner rows of points is a bit smaller.

Your boards seem to be corrugated cardboard ? Choose some realy flat material (white furniture board or something similar) and set the 90°angle carefully.


Yes they are of corrugated cardboard indeed. I cooked up something fast, thought it wouldn't matter as much.
Will correct the points you have mentioned and try again.

Cheers!
karlkohl
 
Posts: 26
Joined: Sun Dec 12, 2010 8:14 pm

Re: Structured light scanning failing??

Postby karlkohl » Wed Dec 15, 2010 5:16 am

MagWeb wrote:Yes, besides the wavy fixing of the patterns ( I mentioned above ) , it seems that the distance of the corner to the inner rows of points is a bit smaller.

Your boards seem to be corrugated cardboard ? Choose some realy flat material (white furniture board or something similar) and set the 90°angle carefully.



Wow! Yes your advice worked!!! Many Thanks.

I had some old glass mirrors lying around, so made a calib corner out of those(of course I used the non reflective side of the mirrors),taped them and at precise 90deg then I ensured that my calib pattern sheets were stuck absolutely flat and perfectly aligned, and the scanner worked on the first go.

I'm stuck at the texture mapping though, you see I have taken images from a DSLR, so what's the process of texture mapping it to the model.
In the 3d scan window, I can only see show camera image/show texture shot, but since I don't have a live cam running I don't have any idea of how to texture map??
karlkohl
 
Posts: 26
Joined: Sun Dec 12, 2010 8:14 pm

Re: Structured light scanning failing??

Postby MagWeb » Wed Dec 15, 2010 8:42 am

try this:
do calibration loading the image shot of an empty corner using the image grabber as cam;
do scanning using the StucturedLightGrabber and load your serie of 12 images with projected patterns (with our without planes);
when you´ve got the scanresult, go back to the calibration page and choose the image grabber as a cam again and load your texture image;
Go to the scanning page and click grab texture;
MagWeb
Moderator
 
Posts: 2499
Joined: Wed Jul 18, 2007 8:48 pm

Re: Structured light scanning failing??

Postby karlkohl » Wed Dec 15, 2010 11:53 am

MagWeb wrote:try this:
do calibration loading the image shot of an empty corner using the image grabber as cam;
do scanning using the StucturedLightGrabber and load your serie of 12 images with projected patterns (with our without planes);
when you´ve got the scanresult, go back to the calibration page and choose the image grabber as a cam again and load your texture image;
Go to the scanning page and click grab texture;



Fantastic!!, that does it, you are simply awesome!!
I just took the image of image02(the one taken with white light) and applied it as a texture to the model, is this the right way or are there any other methods of taking a texture image??
Many thanks again!! Thanks a million!!
karlkohl
 
Posts: 26
Joined: Sun Dec 12, 2010 8:14 pm

Re: Structured light scanning failing??

Postby karlkohl » Wed Dec 15, 2010 5:27 pm

How long do I have to keep the scan going? Is it that with many passes the output improves, at the moment I am pausing the scan after the first pass.

Also I find that there are many holes in the object, the areas that were black in the original object don't seem to appear in the final 3d model and there are other random holes possibly because the Structured light is not bright in those areas.

I noticed that a part of my object had some raised text(about 1 to 1.5mm) from the however in the final 3d model these raised text parts are not seen.

Regards,
Karl
karlkohl
 
Posts: 26
Joined: Sun Dec 12, 2010 8:14 pm


Return to Structured Light Scanning

Who is online

Users browsing this forum: No registered users and 1 guest

cron