Scan of 1-Euro-Coin

[Simon]

Hi,
in diverse discussions here in our forum coins have proven to be a good scanning target for quality verification. So I have also tried to scan a coin. Note: I neither used white light, nor motorized laser movement .

Setup:
- Monochrom firewire camera 1024x768
- Red focusable laser (handheld!)
- High triangulation angle
- No powder or coating

[WalterMo]

Hi Simon,

Nice scan. Very clear.

But if you are using a monochrom camera it's immaterial if a red or a green or white light was radiating.
It seems that the Speckle effect isn't so important.

It would be interesting to see the scan before rendering with 3DSMAX.

[Simon]

Hi Walter,
here is a screenshot of the coin how it is shown in DAVID.

[hal]

Hi,

Impressive result! and without coating... really a good scan. Compliments, to you and to your David.
I think that yours is a 30 fps ccd camera, right? and can you tell me, please, the setup of your camera settings, just for a comparison. I will try, in the next future, my personal test with a coin
Your software broke all its previous limits, everytime

An other question about this amazing scans of coin: from the "upper" movement of the laser (laser held above the camera) to the "down" movement of the laser (laser held under the camera), have you pressed "pause" button or you change the laser direction during the same session of scanning, without any interruption? Thanks in advance.

Bye
Mattia

[Simon]

Hi Mattia,
yes, 30 fps, CCD. I still use the same camera, which is the same as yours. But I changed the lens since my other lens was not capable to capture sharp images in a very short distance. The coin in the camera image was nearly as large as in my first two images above. I have scanned the whole coin in one pass from one position, by just rotating my hand VERY slowly. I set the exposure time to the lowest value possible, so hand tremor played a minor role. And I opened the lens aperture, which gave a better line in my case. And don't forget to bring the camera and laser into optimal focus, where the object (not the calibration pattern) is as sharp as possible.

Could it be true,that different aperture sizes show different intensities of laser speckle noise?
I'm not an expert in optical physics, so maybe one of you can answer this question? An increased aperture results in a lower depth of focus, but this is not a big problem in the case of a coin....

[WalterMo]

Hello Simon,
If your „normal“ lens isn't able to focus on objects nearer as „allowed“ you need not to buy a new lens, you can use C-Mount extension tubes. I think a set would be the best choice.:

http://www.theimagingsource.com/en_US/p ... ics/tubes/


Regarding the speckel noise I have asked a physicist who is working on optics for a long time.

He told me that there is a behavior between noise and aperture size. If the depth of focus can be let small like at scanning a coin it's an advantage to open the aperture to its maximum. A small depth of focus acts out of the image plane like a spline that means like a kind of averaging. So speckels will be smeared.

[Simon]

Thank you very much Walter. That confirms my impression. So you should open the aperture if your object is relative flat.

[florin.topala]

I think we can test the dep of field of camera and after that the obtained value must be copared with the value of depts on the scanning object

[hal]

Hi Simon,
wishing to you, and to all the David team, all the best greetings for CeBIT, I upload an image of my scan of the back side of the 1€ coin.
Isn't pretty as yours scan, but I've scanned it speedy and without a precise calibration sheet.
The coin haven't any coating. Sorry for the texture, but I have a B/W camera.
Bye,
Mattia

[Bongobat]

Hi,

@ Simon and Hal

Are these raw scans or were they smoothed and or interpolated?

Thanks,

Greg

[hal]

Hello dear Greg,
about my coin scans, they are only interpolated (if I remeber well, with a value of 2 or 3), no smoothed. As we can cleary seen, the difference of details between the Simon's front side coin scan and my back side scan, are strictly related with the lens mounted on our CCD camera. In my scan all seems "rounded" because my lens can't do its best job with the object so near to the camera.
Hey Greg, I'm hungry of some others of your great scans Sometimes I surf on to this forum searching some news, but from you and Gunter no new scans... How can I learn more about David, if my favourite Jedi masters of scanning are sleeping (obviously my words are a joke... I see that you and MagWeb are the most precious resources for professional infos and suggestions in the forum, but please, don't forget to scan! Me and others users need to "eat" your scan examples )
I wish you all the best,

Bye, Mattia

p.s.: my previous request to MagWeb and Bongobat to produce more others scan-projects, is extended to all the forum's partecipants: Came on! David need your scans!

[WalterMo]

Hi Mattia,
We cannot „produce“ always good scans!
Like each artist we now and then need a time-out to create new ideas.

A question to your lens regarding very close objects: Do you not use these extension converters?

http://www.theimagingsource.com/en_US/p ... ics/tubes/

Walter

[hal]

Hi Walter,
thanks for the links, no, I never tried these extension converters. But probably I will do...
About my encouragment to produce more scans, it is obviously a joke

You and many others members of this forum, are doing an incredible work of support and development, and that is more than honorable.

But I can refer my own words to myself: from the time that I start a new job, my time is decrease, and as consequence, my scans are decreased.
But I've many projects ready to start and in progress: 2 action figures, 2 car models, 1 little sculpture (same artist of this head viewtopic.php?f=7&t=534 , the last image published) and 1 plaster caster of an ancient sculpture.
And I must to finish the scan of my first car model ( viewtopic.php?f=7&t=790 ).

Time is short, but the night is long and Milena is a patient woman

Bye, Mattia

[Bongobat]

Hi mattia,

about my coin scans, they are only interpolated (if I remeber well, with a value of 2 or 3), no smoothed. As we can cleary seen, the difference of details between the Simon's front side coin scan and my back side scan, are strictly related with the lens mounted on our CCD camera. In my scan all seems "rounded" because my lens can't do its best job with the object so near to the camera.

That was exactly what I was noticing thanks for clearing that up.

@Simon
Sorry, Also could you tell me the model of laser you used for the euro scan?

I am impressed that eveyone seems to get such good results without coating the coins. Is it the camera or the laser I wonder? With both white light and red laser I have bad inaccuracies on an untreated shiny surface. I dont mind coating coins but just curious. If anyone has a theory I would like to hear it.

thanks,
Greg

[hal]

Hello,

@ Bongobat: Probably the "critical factor" to scan without coating is only the camera sensivity: the CCD exposure and contrast values.
But I have, in these last days, such good results with a "filter" for my green laser line: I've putted in front of the lens of my camera a green trasparent sheet (taken from an old glasses for 3D vision - you know... the red and green plastic lenses... ), just for cutting some lenghts of colors and see better only the green light waves.
I know that is better a green filter that have the same green value of color as my green laser, but this is only a raw test.
However this works, and cuts some reflections and color noise (I scan with a normal light condition, not in the darkness).

As visible in the image, with a filter (green for green laser, red for red laser) you can have a better line, more thin and with less noise.
Try it with your webcam and an uncoated coin.

Bye,
Mattia