Page 2 of 3

Re: Calibration Boards

Posted: 21 Jun 2018, 19:04
by mading
Now I have to know why I got 0.18 with original David and the same cameras! :?

I saw you use also a vertical angle. Would you like to share your hardware setup?

I'm looking for a wider Arca style bar: mine is Ok for one camera but a little bit short for two.

Re: Calibration Boards

Posted: 21 Jun 2018, 19:14
by Micr0
martan wrote: 21 Jun 2018, 16:39 Again,

Received the original HP-panels today. First try att the 120mm scale.
The 6 big circles seems much bigger than the prinouts, Is this a factor in getting these (see picture)
error values?
Those look like very good numbers. Let us know if that makes a difference. Please post some pictures of your alignments with the new calibrations if you get the chance.

Re: Calibration Boards

Posted: 21 Jun 2018, 20:28
by 3dxcan
Hp original calibration panel is made of glass or aluminum? I'm wondering how they are printing the dots on the panels accurately too.

Re: Calibration Boards

Posted: 25 Jun 2018, 17:50
by martan
Quite sloppy setup today.

Calibration 120 mm on HP original calibration panel.
Random plastic thing painted with white PlastiDip.

DIY rotary table, 20 scans, 18 degrees. AutoIt-script.
Credit: https://sites.google.com/site/janbeck/3 ... for-david5

Settings, RMS and algnment according to pictures. "Orientation" set to "both"

(Acer P1265 and 2x TIS23UX236 with Computar 12mm MP "Old")

Re: Calibration Boards

Posted: 26 Jun 2018, 06:34
by mading
Thanks for the AutoIt script Martan.
A very nice tools for small objects (or huge turntables :-) ).
I have a question regarding calibration boards: Can you please make a scan of your panels and look for planarity with Geomagic or other sw?
I have the same cameras and the Acer 132 (non-david) and I still have errors of planarity more than 0.2 mm between two different scan of the same panel.

Re: Calibration Boards

Posted: 02 Jul 2018, 16:17
by martan
Quick test today with homemade panel. Milled angles, 3mm glass and just office paper in the laser printer.
I enlarged the 3 circles a bit and this acually decreased the RMS Error to the 0,28-range (from 0,33-0,35)

100mm scale. Then I tried to decrease distance a bit and the error increases with shorter distance (greater camera/proj-angle).

20 scans, RMS in the 0,35-range on the office paper panel. I find tha original HP-panel bulky and it would be nice to have a smaller
hommade panel. Hopefully it will work out in the end. Comments? Good/bad?

Re: Calibration Boards

Posted: 02 Jul 2018, 19:19
by mading
Hi there!
Nice results Martan!
My "woody" 160 panel, printed in a shop with a panel printer.
RMS 0.12-0.29 (why? maybe not properly focused..)
Set-up is vertical-horizontal, now I got better surfaces without waves! Thanks to BigBomber.

-> Setup is 'inner corner'
- RMS error is ok: 0.296544[..]
Horizontal open angle: 25.0155
Vertical open angle: 15.6929
-> Setup is 'inner corner'
- RMS error is ok: 0.127285[..]
Horizontal open angle: 24.9592
Vertical open angle: 15.7101
Angle between calibration panels: 90 degree
Optimizing Multi-Camera Calibration...
Calibrate Projector.
Angle between calibration panels: 90 degree
Size of camera's center pixel at background plane: 0.2536 mm

It's a fast try, with projector just sitting on a wood platform. Not even fastened to it.
Even in this situation I got better surface quality.

Comparison between two stereo scans at 0.08 mm:
2018-07-02_20-57-43.png
Two test scans (one flipped 180° to the others):
2018-07-02_21-01-45.png
Panels:
2018-07-02_20-50-07.png
2018-07-02_20-50-09.png

Re: Calibration Boards

Posted: 02 Jul 2018, 21:48
by Micr0
are you using 2 cameras? One over the projector one under?
mading wrote: 02 Jul 2018, 19:19 Hi there!
Nice results Martan!
My "woody" 160 panel, printed in a shop with a panel printer.
RMS 0.12-0.29 (why? maybe not properly focused..)
Set-up is vertical-horizontal, now I got better surfaces without waves! Thanks to BigBomber.

-> Setup is 'inner corner'
- RMS error is ok: 0.296544[..]
Horizontal open angle: 25.0155
Vertical open angle: 15.6929
-> Setup is 'inner corner'
- RMS error is ok: 0.127285[..]
Horizontal open angle: 24.9592
Vertical open angle: 15.7101
Angle between calibration panels: 90 degree
Optimizing Multi-Camera Calibration...
Calibrate Projector.
Angle between calibration panels: 90 degree
Size of camera's center pixel at background plane: 0.2536 mm

It's a fast try, with projector just sitting on a wood platform. Not even fastened to it.
Even in this situation I got better surface quality.

Comparison between two stereo scans at 0.08 mm:
2018-07-02_20-57-43.png

Two test scans (one flipped 180° to the others):
2018-07-02_21-01-45.png

Panels:
2018-07-02_20-50-07.png
2018-07-02_20-50-09.png

Re: Calibration Boards

Posted: 03 Jul 2018, 04:19
by martan
Mading,

What happens if you turn the panel just a little clockwise?
Looks like that panel picture #2 is the camera with higher RMS Error. Correct?


M

Re: Calibration Boards

Posted: 03 Jul 2018, 05:51
by mading
Both cameras under the projector.
Sorry I had to move everything from my bench, so I'll try to check which is which (cameras).

I would like to make a gig for tilting the projector according to panel size, in order to use it with smaller panels (120-60).
Actually in the horizontal setup I'm limited to the 60 panels, due to K132 lens position.
It would be nice to make stereo at 30.. I still have to try the mod for the focus lever (courtesy of Avogra).