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

DAVID 5 Beta

Discussions about latest DAVID developments, beta versions...

Re: DAVID 5 Beta

Postby avogra » Sat Jul 30, 2016 10:55 am

You are right, the textured scan was before 5.0.3. I will try it again.
Pattern orientation "both" will probably double scan time, right? then again, I think a camera arrangement left + top might make it easier for some objects to see into every spot by both cams and the projector. Your suggestion sounds great to. I just have to make my rig more flexible :roll: 8)
Projector: Acer K11 with Achromat f=400mm | Cameras: 2x Logitech 9000 Pro with custom C-Mount body | Lenses: 1x Computar M1214-MP, 1x Computar M5018-MP, 1x Schneider-Kreuznach Cinegon CNG 1.4/12
avogra
 
Posts: 23
Joined: Sat Apr 30, 2016 1:50 pm

Re: DAVID 5 Beta

Postby micr0 » Sat Jul 30, 2016 6:10 pm

Hi Sven,
I finally got a chance to put some real seat time with the new version (BTW David 5.0.3 solved my calibration issues). My first impression is that the individual scans are right off the bat great. better than David 4 by a significant amount. I'm getting even better detail and more of the object is scanning reliably. That is that I can lower the outlier and quality check way down (or even off) and still get very clean scans. It also seems to capture more of the extreme acute angle detail as well as getting deeper into negative space. However I too am having some of the issues that Walter seems to be having, though to a lesser extent. I have allowed my projector to warm up fully (on for hours) and have been vary careful in my calibration procedure yet still getting some distortion that is causing scans to appear as though they are not aligning correctly. I have in the past traced this to lens aberration at the edges of the image and refractive distortion caused by the aperture at hi F-stops (18-22). With the exact same setup I'm using now I was getting accuracy of .05 - .07% consistently. Now with David 5 I've measured deviation of about .2%. I'll have time to work on it again today and I will be adding a second camera and see if that improves things. If I can tame the distortion I'd say this is a great upgrade. Well worth the price (I still believe it shouldn't have been 1.8x the difference in version prices but I've said my peace on that issue) Thanks.

Some questions/notes.

David still really needs a White light button for setups so we can better see what of our subject our camera is viewing (its sometimes hard to see with the patterns). This is not the same as turning on the room lights for those of us who scan in a dark room as I also want to be able to see where the features of the scan subject may be blocking the projected image. This too can be difficult to see with the line patterns. I also don't want to have to keep turning the room lights on and off just to adjust a model when I'm trying to capture some missed detail at the end of a series of rotations.

Are the "auto. add to list" and "auto. align to previous" buttons coming back? Or did you just hid them somewhere new?

Its nice to be able to specify the scan name and number in scanning mode, but the scan number counter doesn't reset when you close the program or start a new project. I opened David the next the next day and started scanning and my first scan was #43.

BTW I got to play with some $60K state of the art scanning systems last week and in a nutshell, you should be very proud of David 5!!
µ
micr0
 
Posts: 315
Joined: Wed Mar 19, 2014 3:54 pm
Location: NYC

Re: DAVID 5 Beta

Postby cuongtran » Sat Jul 30, 2016 10:51 pm

hello there!
i've been trying to calibrate 2 camera with david 5.0 but can not.my camera are usb 2.0 pixelink e533 3mp.
the first problem is the resolution of 2 camera differrent.i can not make a change on the camera format.and the camera properties dont show up.for example :1 is 1600x1200 and the other is 2048x1536px.due to the differrent solution and framerate,i can not calibrate (differrent solution will effect to the distance between the calibration to the camera and projector).2 camera can show up good
the second problem is after i change the other camera so that 2 camera got the same solution.1 camera show up good but the other camera is not good.some time it show up and some time don't show up.both camera can not change the camera format and camera properties.
all the function on camera properties are hide (exposure,brightness,gamma....)
does anyone know what is the problem and how can i fix that problem???
thanks!!!
cuongtran
 
Posts: 8
Joined: Sun May 22, 2016 6:10 pm

Re: DAVID 5 Beta

Postby hal » Sun Jul 31, 2016 10:13 am

Hello dear Sven,

thanks for your kind reply. I've used two projectors, with same results: the David's SLS1 Acer and the Acer one from David SLS2. But obviously something was wrong during my quick tests. I need to try again, then I will report. About temperature, the results didn't change in relation that the prjector is turned on from 5 minutes or 1 hour. The distances from scans is around 5 mm. Means there is a problem in my configuration. I will work on it :roll:

Thanks and best regards,
Mattia

p.s.: I agree with micr0. A "white light" projection feature will be great, because it will help a lot during the calibration steps.
User avatar
hal
Moderator
 
Posts: 1527
Joined: Thu Mar 01, 2007 4:08 pm
Location: Cavaria con Premezzo (VA), Italy.

Re: DAVID 5 Beta

Postby micr0 » Sun Jul 31, 2016 5:17 pm

I got the whole day to work with David 5 and add a camera to my set up. And i spent a lot of time trying to get good results. The scans were clear but they were not aligning and I was having the exact problem Hal is having. I'll explane:

This is a picture of a Scan I did with the Imagingsource HD cam (54fps) with david 5. LG HD projector. Camera to the left of the projector and calibrated with 300mm panels:

Dav5 trans 1.jpg

I added a second camera, which was a David cam 3M which is 25fps and 1280x960. I can see immediately that though David will work with both cameras it doesn't really like two different cameras at two different resolutions and frame rates. I had no problem calibrating both cameras with the 300mm calibration panel and both cameras set at about 16deg to the projector (36deg included). However when I did my first scan which was of the calibration panels I could see how off the 2 scans where (FWIW I was hoping that David combined the data from both cameras immediately to produce a single, more accurate and or detailed scan).

Dav5 2cam cal panel.jpg

Dav5 2cam Deviation.jpg

You can see that the measurement is 3.5mm deviation between the 2 scans. The measurement is taken at ~180mm behind the focal point of the projector/camera (setup is focused best at the second ring about 135mm from the corner).
µ
micr0
 
Posts: 315
Joined: Wed Mar 19, 2014 3:54 pm
Location: NYC

Re: DAVID 5 Beta

Postby micr0 » Sun Jul 31, 2016 5:23 pm

When I replaces the calibration panels with the transmission housing I got scans that looked like this:

Dav5 2cam Deviation and Fuzz.jpg

Dav5 2cam Deviation and Fuzz2.jpg

The fuzz is back at the edges and the scans do not match. I tried different camera angles, calibrating each camera individually, different exposures, different calibration panels (my calibration panels are glass held together by mounts I made on a CNC mill and measure 90.00156deg as measured on our CMM). I don't know what I'm doing wrong, or is it that David doesn't process two different cameras well?
µ
micr0
 
Posts: 315
Joined: Wed Mar 19, 2014 3:54 pm
Location: NYC

Re: DAVID 5 Beta

Postby hal » Sun Jul 31, 2016 5:27 pm

... interesting. exactly the same issue. I thought I was missing something in my workflow because didn't played long time with David 5... but as far as I can see from your tests, there is a mystery :shock:
Hope next week will have time to try more, here.

Thanks for sharing, micro0!

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

Re: DAVID 5 Beta

Postby Curiousjeff » Tue Aug 02, 2016 10:21 am

Hi Sven,

You wrote:

"That is the fastest and most comfortable solution, and it makes sure that the texture fits perfectly on the scan.
In your application, why do you need to grab the texture separately? "

I usually don't use texture myself, but the fact of taking the texture separately was specifically asked by the doctors at the hospital when I made the presentation. They are dermatologists.

During the "texture photo moment", they wanted to be able to put on more lights then just the one coming from the projector, for example with two banks of lights coming at 45 degree at the left and the right of the patients face.

The demonstration was done with David 4.

There was no way to trigger automatically these extra lights "on" as far as I know just when the texture is being grabbed, but I told them that they could take the texture photo after the scan which would allow them to put on these extra lights on at that moment.

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: DAVID 5 Beta

Postby Sven » Wed Aug 03, 2016 8:39 am

Hi Micr0,

micr0 wrote:David still really needs a White light button for setups so we can better see what of our subject our camera is viewing
I don't quite understand... Can you please show an example live image?

micr0 wrote:Are the "auto. add to list" and "auto. align to previous" buttons coming back? Or did you just hid them somewhere new?
That depends on the Scan Mode you select. Single Scans are not auto-added. "Manual Scan Sequence" is what you need, it will add all scans and align them after the sequence is done. Also "Automatic Turntable" does the same.
When you try that, check "Auto. Next Scan" --> DAVID will watch the camera image for movements and delay the next scan until no more motion is visible.

micr0 wrote:Its nice to be able to specify the scan name and number in scanning mode, but the scan number counter doesn't reset when you close the program or start a new project. I opened David the next the next day and started scanning and my first scan was #43.
That is deliberate. Why not continue yesterday's scan sequence? It's easier to set it back to 1 than to find out how many scan were already done. Also it's worse and confusing to have several "scan_01" in a project - better have scans with large numbers.

Sven
User avatar
Sven
DAVID Developer
 
Posts: 1541
Joined: Mon Jan 22, 2007 8:59 am
Location: Braunschweig, Germany

Re: DAVID 5 Beta

Postby Sven » Wed Aug 03, 2016 8:43 am

About Stereo with different resolutions:
If the cameras have DIFFERENT resolutions, DAVID5 seems to create distortions in some cases. Although we definitely recommend to use two cameras with the same resolution, it should be possible to work with different cameras.
We are working on this.
Sven
User avatar
Sven
DAVID Developer
 
Posts: 1541
Joined: Mon Jan 22, 2007 8:59 am
Location: Braunschweig, Germany

Re: DAVID 5 Beta

Postby Sven » Wed Aug 03, 2016 8:50 am

Hi Jeff,

Curiousjeff wrote:During the "texture photo moment", they wanted to be able to put on more lights then just the one coming from the projector, for example with two banks of lights coming at 45 degree at the left and the right of the patients face.

The demonstration was done with David 4.

There was no way to trigger automatically these extra lights "on" as far as I know just when the texture is being grabbed, but I told them that they could take the texture photo after the scan which would allow them to put on these extra lights on at that moment.


When you take the texture with a monochrome camera, usually it won't help to add constant light. The red-green-blue-black-projection method we use removes all external lights from the images (based on the "black" projection).
In that case, you can deactivate Advanced Settings -> Texturing -> ColorTexturesWithMonochromeCameras, then you will get only a greyscale texture. In DAVID4, then the additional light will help. For DAVID5, we will consider an extra "Texture" button.

Sven
User avatar
Sven
DAVID Developer
 
Posts: 1541
Joined: Mon Jan 22, 2007 8:59 am
Location: Braunschweig, Germany

Re: DAVID 5 Beta

Postby micr0 » Thu Aug 04, 2016 3:42 am

Sven wrote:Hi Micr0,

micr0 wrote:David still really needs a White light button for setups so we can better see what of our subject our camera is viewing
I don't quite understand... Can you please show an example live image?


Hi Sven,

A white light button on the set up page of David would just have the projector projecting white light. No pattern. Or turn the pattern off for when one is positioning the camera and subject. I scan in a dark room and when I'm looking at the live camera image to see what is in the cameras view it would be useful to turn off the strip pattern and just have white light. The stripped pattern can often make it hard to see detail and to tell the difference between a dark stripe and a shadow or negative space during set up.

Sven wrote:
micr0 wrote:Are the "auto. add to list" and "auto. align to previous" buttons coming back? Or did you just hid them somewhere new?
That depends on the Scan Mode you select. Single Scans are not auto-added. "Manual Scan Sequence" is what you need, it will add all scans and align them after the sequence is done. Also "Automatic Turntable" does the same.
When you try that, check "Auto. Next Scan" --> DAVID will watch the camera image for movements and delay the next scan until no more motion is visible.

micr0 wrote:Its nice to be able to specify the scan name and number in scanning mode, but the scan number counter doesn't reset when you close the program or start a new project. I opened David the next the next day and started scanning and my first scan was #43.
That is deliberate. Why not continue yesterday's scan sequence? It's easier to set it back to 1 than to find out how many scan were already done. Also it's worse and confusing to have several "scan_01" in a project - better have scans with large numbers.

Sven

I understand your logic but I have never had duplicate scan numbers in David 4 and Now in David 5 I have had duplicates more than once. Wouldn't logic suggest that the software would just assign the next higher scan number to a subsequent scan when resuming a project? I only say something because this actually adds a step to the work flow as now I have to make sure I manually check the scan number each time I start a new scan lest I get duplicates, or out of sequence scans.

The above, however is relatively unimportant. What is important is I can't seem to get proper aligning scans in David5. I had casting I had to scan today and couldn't get good alignment (un-distorted scans) in David5. I was only using one camera (not two) and tried all sorts of calibration and set up variations and no luck. I went back to David4 and with the same set up and calibration got great results. In David 4 I got a bit more edge roughness but had no difficulty getting the 12 scans of one part rotation to align well. Try as I might I couldn't get the same results in David5.

I wish I could post pictures, but this was client work and not mine to post. If you need images I'll try post one of my projects to show you what I mean.
µ
micr0
 
Posts: 315
Joined: Wed Mar 19, 2014 3:54 pm
Location: NYC

Re: DAVID 5 Beta

Postby Sven » Thu Aug 04, 2016 11:40 am

Hi Micr0,

Hmmm it seems I don't understand your workflow. As long as you don't manually change the sequence number in DAVID5, you will never have duplicate numbers again. And when you reset to 1 for every new scan project, you won't have duplicates inside your projects either...

About the distortions: That's strange, we'll try to reproduce. We didn't have any such problems during DAVID5 testing. Is there anything special about your set-up?

Sven

PS: Unfortunately I'll be out of office the next 2 weeks.
User avatar
Sven
DAVID Developer
 
Posts: 1541
Joined: Mon Jan 22, 2007 8:59 am
Location: Braunschweig, Germany

Re: DAVID 5 Beta

Postby micr0 » Thu Aug 04, 2016 2:06 pm

Hi Sven,

My set up should be pretty straight forward. LG HD LED projector, Imaging source monochrome1920X1200 camera (global shutter). Camera to the left (as one stands behind the camera looking at the subject) of the projector and calibrating at ~21deg. Projecting slightly downward to 300mm calibration panels in a negative (outside) corner configuration. The subject being scanned (a sand cast aluminum engine cover) is about 500mm long 250mm wide and 200mm tall.

I may get time today to try to document more completely my setup and problems and I'll post here when I do.

Thanks

P.S.
I also was thinking of trying a setup with the camera mounted above the projector which is how I used to have it set up over a year a go. I change to having the camera to the side when I got the HD camera and projector because the aspect ration of the HD equipment. Thoughts?
µ
micr0
 
Posts: 315
Joined: Wed Mar 19, 2014 3:54 pm
Location: NYC

Re: DAVID 5 Beta

Postby hal » Tue Aug 09, 2016 10:32 am

Hello Sven,
here an other test.
I try on an other computer and the video streaming works correctly (that means my other computer have an issue with USB and TheImagingSource camera).

Two cameras at same resolution 1280 x 960 25 Hz Y800.
Calibration works correctly.

Immagine.jpg

But the result is the same as before.

2016-08-09_11-12-16.jpg
2016-08-09_11-12-21.jpg

The distances between Extended mode scans (blue and green one) is huge and the Stereo mode scan (yellow) is calculated in the middle.

2016-08-09_11-11-55.jpg

The Stereo mode works fine and reduce drastically the noise, but the Extended mode still giving me troubles.
The resolutions are the same, but the lenses, that are both 12 mm, are quite different. So I guess that the main problem could be the correction of the optical deformation that generate the issue. I guess David 5 need same hardware, also same optics in order to work correctly. Right?

Thanks and best regards
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 2 guests

cron