For Watson & Holmes

[MagWeb]

Hi,

just a little tip

scan at 960x740 , with PRO 9000, using the M3 microprojector + added lens, scale 12mm

Gunter

[leo]

Thanks for sharing. As usual very impressive.

Leo

[Bongobat]

Hi Gunter,

WOW

I think you are the real king of micro scanning

Greg

[hal]

WoW

Gunter, or you have a very big fingers or your precision is unbelievable!
Thanks for this scan test: this reset the bar!!!

Bye, Mattia

[benoy]

Gunter great work. How did you get this kind of precision? Can you share the magic about your scan tech? Please add some pictures about your scan set up also.

Great work

Regards
Benoy

[MagWeb]

Hi,

I did this scan with a projector. When you focus a projector on a short distance, you get a much thinner line than a laser. The use of a projector seems to guarantee a more real-time communication DAVID/COM/Projector.
So I use one line progress/frame in the projector. I do NOT use the remembered planes using a projector. Using the planeless feature of DAVID, I can focus the projector on the object without need to have a sharp line on the planes too. The depth of sharpness is a main problem of such a microsetup.
I added a image of the used setup and a 360° scan of one of these HO model train sculptures: Tommy s Uncle Ernie

Gunter

[Bongobat]

Hi Gunter,

That is strange I have had a setup that looks exactly like that in my head except for a few differences.

1. instead of the regular corner you have there I was thinking of using first surface mirrors for an easier 360 scan but in your setup here I think your corner is only their for a reference scan right?

2. My projector is not LCD but old fashinoned razor slide and I was thinking of adding a linear actuator attatched to the slide to move the line vertical as required for the mirror scanning.

My question is have you tried the mirror technique with your LCD projector and does it work?

Alternatively you could of course add a tiny stepper turntable to move the small items and not worry about mirrors. Since everything could be automated it would just be a longer wait which is what I will problaby end up doing.

Greg

[MagWeb]

Hi Greg,

You´re right, corner and patterns are only there for calibration and reference scan.

I´m not sure whether scanning with mirrors will give good results at this dimensions.

Generally, I prefer to use the mirrors only when there´s no other way to get 360°, for the reflected meshes are always a bit worse than the ones grabbed directly (or you are satisfied with a rather rough scan). E.g fragile objects, that may change their topology if you´d move them..So scanning insects could be such a case

Using the technic at such a small size, I see a main problem focusing projector and cam. For both there are two different distances:
A: Projector(cam)/object and
B: projector(cam)/mirror/object

My optics of cam and projector give a rather short field where you can set the focus right (maybe other optics may give better results). E.g.: I can get a sharp line either on the outer edges of the corner or right in the corner.....

Another problem will be that you´ve to cover the mirrors for calibration/reference scan. So you will have to have a very precisely fitting second corner in front of the mirrors, that will be removed after a reference scan. For in that dimension a 1/10 mm is quite important I guess there maybe a problem of precession.

Gunter

[RAYA]

Hi Greg

You wrote that :

adding a linear actuator attatched to the slide to move the line vertical

Does that mean you are going to move slide to have linear motion instead of rotation ?

Raya

[Bongobat]

Hi Raya,

Yes, It was an idea I had because I discovered I already owned a neat little bipolar linear actuator I originally was thinking of using it to move a structured light pattern but thought maybe it could be used with Gunter's mirror technique to ease scanning an insect. However since Gunter says that you can never get as good of a result from the mirrors as from the normal way and I am wanting high precision I think I will just make a small turntable or something similar.

Greg

[Masteron]

Hi Gunter,

this is some impressive work you did there! Regarding 3D-scans of a human finger I have never seen such a high level of detail.
I really hope you continue this approach.
But, just out of curiosity: have you read the post of Simon regarding the use of structured light (viewtopic =p7234).
Since you already use a projector (a great idea by the way) this should (?) be relatively easy to try out.
It would be quite interesting how accurate and fast this new approach is in comparison to the standard method.
Additionally I have one favor to ask of you:
I am particularly interested in high resolution 3D-models of human finger pads - would it be possible for you to send me a sample modell?

cheers,

Ralph

[Tirdy]

Hello

After testing my present setup for micro scanning, with a focusable red line (5mW) Laser, and merely dissatisfying results, I reach the decision to use a pico-beamer instead of the laser...I compared several beamers and think the MPro110 from 3M is the best solution, with a minimal picture diagonal of 120mm (@0,21m distance). Although the luminosity of the MPro120 (13lm) is almost twice as from the MPro110 (6lm), the minimal picture diagonal is just 200mm...

@Gunter:
You added a lense to Your MPro110, what was the main reason for it? With a solution of 640x480 @ 120mm you still can get a laserthickness of 0,15mm. Or is it not sharp enough with that settings? What kind of lense have you used?

Thanks in advance..I hope I can submit some nice first scan results soon

[MagWeb]

Via mobile

lense from an old Photo optics

i condensed the projected image to the size of a stamp. It is best to put all the possible res onto the small object

[Tirdy]

Hello,

today received my MPro110, but I didnt get the first results I was lookin for, just a few more questions :/

1. What is the sense of the 3 choice-buttons : Red Laser, Green Laser, Small Shadowline? Is this important for David calculating the thickness of the line? What shall I check when using a Projector-Line? I didn't recognize any differences..

2. I want to condense my projected image with a mounted lense, like Gunter did it. But im a layman in optics and I have no idea what kind of lense I have to use? Can u please help me?

Thanks alot

[MagWeb]

Hi,

ad 1. The color options only make a big difference if your cam is set on colored mode (saturation > 0). In case of a red laser DAVID ignores the green and blue channel. In case of green laser the red and blue channel. In case of thin shadow line it should detect all channels. But as said before, set on BW ....
I prefer the thin shadow line (which works also with a white line on black).

Scanning with a projector, it is most important to start the scan on a totally "black" projection (which always will show some light, for the LCD inside the projector isn´t able to suck all the light where it should be black). DAVID, in its default setting, defines this first image as not being hit by the beam and subtracts this image from all others.

ad 2. Sorry I´m no master of optics too. I simply hacked an old photo SLR camera(canon "AE 1 Program") standard optic. It gave three collecting lenses. The two bigger ones, put in front of the MPro gave the size I wished to project. The smaller one fitted to the cam to make it less wide angled.

A problem will be the accuracy of glued patterns in this setup. I engraved them onto the boards before cutting them on my CNC. If there´s some problem PM me..

Gunter