The Atomium

[WalterMo]

I think to scan an object like the Atomium is a good job for a turntable.

The original is standing in Brussels and was built for the World's Fair in 1958.
http://en.wikipedia.org/wiki/Atomium

Mine was built by me of beechwood and is about 21cm tall. It was mounted at the stepping motor shaft without the table disc. The Atomium was scanned at eight table positions à 45°. Using the Align function of DAVID „Rotation around y-axis“ by setting the used angle of the scan it was very easy to put the eight automatically together.
The fusion was made by Poisson at a resolution of 250.

Walter

Edit: The scans were made with stand-alone devices. That means the impulses for the stepper motor for the laser and the turntable didn't come from DAVID via the COM port.

[Eric]

Hi Walter,

This look really great! Nice job

Cheers,
Eric

[hal]

Hi Walter,

nice scan! seems that the turntable setup works perfectly. Very precise and clean.
The next step is scanning a real unit cell of an iron crystal, but not magnified 165 billion times. Your next proof is try to scan it as is, in real size, using an atomic microscope!

Compliments for the scan,
Mattia

[WalterMo]

The attached screen dump shall show how imperfect the object was scanned with only eight single scans. I had used my outdoor rig and the position of the laser was still above the camera. That means that the spheres could only be illuminated and seen from the camera at their sides.

But the Poisson fusion was so „intelligent“ to close the gaps nearly perfectly. Especially because the gaps were sphere gaps.

In the meantime I have put the Atomium in the WIKI Multimedia Gallery:
http://www.david-laserscanner.com/wiki/ ... tomium.pdf

Walter

[WalterMo]

Here is another scan made with the turntable at 8 positions à 45°. Because I wanted to grab the exture I equipped the rig with two neon lamps which give a diffuse light to the object's surface. For the lamps I chose 12V types to be independent from the mains for future outdoor scans.

I am not completely satisfied because of the not very clearly readable letters. The eight texture photos are of far better quality.

But I must say this scan was easy to do because the laser needs only one position (above the cam) and the alignment of the single scans was made by DAVID only telling it the used angles.

As a pdf-file without the texture you can find the lamp in the WIKI Multimedia Gallery:
http://www.david-laserscanner.com/wiki/ ... o/lamp.pdf

Walter

[Bongobat]

Hi Walter,

I just looked at your pdf. Very nice. Very clean fusion and that is very hard to do sometimes with such a symmetrical object. Nice atomium too much better than an upsidedown eifel tower

You can include textures in pdf files you just have to make sure you choose your texture file in the export options (meshlab) I think Daz does it automatically. Perhaps you just wanted to show the untextured mesh though?

Greg

[hal]

Hi Walter,

Very nice scan! I imagine thet, as the Disney character Gyro Gearloose (Archimede Pitagorica in italy), this last scan is your "Little Helper" and you sometimes exclaim: "Eureka!!!".
(http://duckman.pettho.com/characters/gyro.html)
Keep your great scan-work lighted,
Mattia

[WalterMo]

Hi Greg and Mattia,
Thanks for your comments.
You shall know I am a bit lazy and therefore I prefer simple scanning, e.g. by using a turntable.

But of course I will try to find an easy way also to show „moveable“ scans with a texture on them.
Perhaps by using the way you (Greg) mentioned or by the new flash-translation of Simon and Sven.

Yes, I know our German „Gyro Gearloose“ named Daniel Düsentrieb very well. He is my inspiring example. Most of my ideas on this forum came from him. Sorry, I have forgotten to mention each time the real (his) source.

Walter

[WalterMo]

It's difficult to reduce the number of faces of a spherical object for a nice presentation as a Flash. Therefore the bulb isn't perfectly round but has some corners. But the lettering is nearly OK. The hole at the top is a lack due to the missing texture shot from above. The bulb wasn't blown out like an egg!

Walter

[hal]

Great! and good texturing

[WalterMo]

Hi,
This is my third scan with the turntable. The used set-up was the same as with the bulb above.

But this time I used two axis for the rotation of the object:
A) The perpendicular axis going through the ground plane of the car.
B) The longitudinal axis of the car.

Each scan of these two groups was made at 8 positions á 45° and then stitched by DAVID by its align function „Rotation around y-axis“ by the statement of the used known angle.

Afterwards the groups A and B were stiched by the manual alignment. For this job the ends of the four flat springs of the Opel were a good help (and of course other strikingly spots).

The fusion was made by Poisson with at resolution of 610.

You can also see the veteran car as a pdf in the DAVID Multimedia Gallery:
http://www.david-laserscanner.com/wiki/ ... el1909.pdf

Walter

Edit: The 3-stereo image was made with the free software from here:
http://www.stereoeye.jp/software/index_e.html
as suggested here by Greg:
viewtopic.php?f=2&t=1404&start=15

[Khalid Khattak]

Walter ..this is a great scan.. can you tell us about the scan resolution , can you post the actual picture of that car.. thanks

[WalterMo]

The resolution of the modified Logitech Pro 9000 (now Pro9001):
viewtopic.php?p=7805#p7805
was 960x720 pixels at 15fps.

Only one time swept with the laser line over the car at each turntable position.
Stepping motor scanner from here:
viewtopic.php?p=7935#p7935
Single scans were 1x interpolated, not smoothed.

Calibration corner with 28 mm scale. Green 5 mW focusable laser from the DAVID Shop.
Can you perceive that the right side footboard is rusted through? Better to see at the pdf. No surprise, it's an old car.

A photo of the car (8 cm long) you can find here:
viewtopic.php?p=8198#p8198

and from the original car:
http://www.opel-specials.de/opel-doktorwagen.htm

If you intent to visit Germany with your car, don't be afraid, all of our cars have meanwhile the steering weel on the left side and we are driving on the right side of the street.

[WalterMo]

Some time ago Ordibble Plop told us that this shell comes from his homeland New Zealand:
viewtopic.php?f=6&t=90&p=336&hilit=shell#p336

Now I have scanned it again using the turntable and two stepping motor driven lasers. One laser was standing above and the other below the camera. By using two lasers it wasn't necessary to change anything of the set-up because all parts of the frontside of the spherical object could be illuminated, also the shadowing sides of the pimples. So the scanning process could be solved in a very short time only by sitting in front of the PC, controlling some switches.

Some details:
Two identical focussable red lasers from the DAVID Shop.
Modified Logitech Pro9000 (now Pro9001) at 960x720 pixels @ 15fps.
Diameter of the shell: 65mm.
Eight turntable positions á 45°.
Each of the eight scans was 1x interpolated but not smoothed.

The two motors were alternately driven from one controller / motor driver. Only by toggling one three-way switch for the lasers and the motor coils.

Walter

Edit: Now you can see the shell as a pdf in the WIKI Multimedia Gallery. The original has big holes at the upper side and underneath. They were closed by the Poisson fusion because I had "forgotten" to scan the holes.
http://www.david-laserscanner.com/wiki/ ... /shell.pdf

[WalterMo]

I have tried to present the shell with texture as a Flash. It wasn't very easy because the original obj file, fused with Poisson with a resolution of 500, had a volume of 76MB. This to reduce to a limit of 2MB, allowed to post to our forum, wasn't possible.

So I fused the eight single scans again with a Poisson resolution of only 120 (4.55MB) and reduced it to 4% (1.71MB).

Not too bad for a spherical object with texture.

Walter

P.S.: Do you know a possibility to suppress the photo of the eight texture shots?