Outdoor scanning

[Ordibble Plop]

But isn't the background ignored via a 'tolerance' setting so wouldn't be taken into account?

I think your theory is logical, although would speculate that the problem occurs at the edge of the lettering, i.e. where the width of the laser line is 'half' on the bright lettering and half on the duller PCB. This might then skew where DAVID calculates the middle of the line because the line is no longer of consistent brightness over its width or at least the brightness over its width is not symmetrical. Or does DAVID have a built in tolerance for correcting for this?

Edit: hehe, but thinking about it more, if this were the case it would occur in only one dimension, which doesn't seem to be the case, i.e. both the vertical and horizontal edges of the letters are raised.

Interestingly though, the dark lettering on the white sticker in the bottom right corner appears to be recessed, so there is definitely something going on that appears to be related to the brightness of the line (or maybe relative reflectance or light scattering?). And the white lettering on the black component at top centre-left is not raised - has this been flattened during fusion while the other lettering hasn't? Have we seen something similar before when the callibration pattern with its dark spots has been scanned?

[cnc Dave]

This is looking really promising Walter, I know it's well documented but I still haven't quite got it clear...

You do a reference scan with the calibration chart as usual, (as seen in your last photo) and then you can scan at any distance your laser can fall on?? So you have to re-focus the camera? You did your scan with one slow sweep?

What happens if you scan off the edge of your wall... do you get an image of the doors behind and the edge of the wall?

I'm thinking of building a similar set up but just using a very geared down motor. With a cam and micro switches to turn off the laser and auto reverse the motor back to the start. I was thinking of having a shield infront of the laser so as it comes up it the light will come over the shield and always start in the same place. Very simple to adjust the shield and the start point. Or maybe have another micro switch and a relay to turn on the laser.

I pointed my laser out of the window one night and it put it's line right across 3 gardens and could light up trees quite some distance away. Does that mean you could scan a whole house in one go :0)

I'm building a large CNC router and I have a polyurethane injection molding machine (which I made). You could send me a file of you house Walter and I could reproduce it as an accurate plastic model.. or even build it full size!

I could go down to the local Aston Martin garage one night and come back and build an accurate Aston Martin body kit for my car :0P The possibilites are endless!

Did you know the next best thing to DAVID is £17,000 and it can't scan cars because it can't scan metal. DAVID will have to be renamed soon, because it is truly a gaint amounst gaints!

[cnc Dave]

I wonder what Aston Martin would think if they come in one morning and one of their cars is covered in talcum powder

[WalterMo]

I think the reason that the white letters on the circuit board are raised is due to the material of the board itself. It's made of paxolin or epoxy, both are a bit translucent. You can see it if you hold such a board against light. Only the parts where the copper on the backside wasn't etched are
completely opaque. Therefore some laserlight can infringe into the board and isn't reflected just at the surface where the letters are printed. But this is only an idea. Perhaps I am completely wrong. We should check it with different materials.

Thanks cnc Dave for your offer to recreate my house of polyurethane with your 3D plotter. But one house is sufficient for me.
Back to my outdoor scan:
Yes, I did the reference scan as usual and was then able to scan everything which was in front of the camera. This isn't quite correct of course: The camera and the laser were focused to a target about 50 cm away. Objects more fare away wouldn't be adapted so precisely due to the depth of focus of the camera and the laser. But more important is the fact that objects which are fare away aren't be illuminated by the laser at all! Think of the intersection angle of about 30°. The calibration corner should have circa the same size as the object.

It isn't allowed to re-focus the camera after the reference scan!

I had used the Logitech Pro 9000 which has a very wide-angle lens. This kind of lenses have the feature to have a large depth of focus. That means if I would have pivoted the tripod head to the frontdoor I could have scanned it if the laser beam would have hit it (in the camera image). Not pin sharp of course. The lower light back reflection of objects more fare away could be easily compensated after the reference scan in the camera settings.

Yes, the nameplate was scanned with one slow sweep.

Mattia, the intention to scan a complete car is a great thing. In the true sence of the word. But you are an Italian, why not to start with a Fiat ?
We are very curious and cross our fingers for a fine weather in Milano.

Walter

[joecnc2006]

What would be great is an instructional PDF showing the steps for a beginner to get into scanning without the panels (and the standard David). showing how to setup David in the settings, with pictures, and taking you through the whole process to a final scan like a step by step, it is difficult to try to figure it out from bits and pieces from the Wiki and the forum. Anyone up for the challenge? It would be worth paying a small fee for it.

[WalterMo]

Hi Joe,
Yes, it would be nice to have something like a how-to-do intruction for outdoor scanning. But up to now all the developements for this new method are in progress.
When I am thinking to my own situation: OK, perhaps I am one step ahead than some of you, but also for me is everything something like new ground. I always try to explain what I have just done and how I have made it.
This isn't limited to me of course, also the other „specialists“ of this forum bare how they had achieved something.
Particularly MagWeb has written more than one detailed article on the Wiki.
Even for Simon and Sven it would be difficult to give us a recipe. As I already said: All features of DAVID are in progress. But I am sure, if a certain level of outdoor scanning is reached we will post a summary.

[hal]

Hello,
@ Walter: about your latest question to me, two post ago: I haven't a FIAT car, sorry. The FIAT 500 will be a great thing to scan, and this car is a symbol, a kind of icon, here in Italy. Unfortunately I haven't this car model available for scanning.
The only FIAT that I've scanned is a 1:43 model of FIAT 127: viewtopic.php?f=7&t=790&p=6690#p6690 .
I hope that you like it

Best regards,
Mattia

[WalterMo]

This is my second outdoor scan, also in front of my home. That means I could still use the mains to supply all electronic devices with current.
A special reason for the scans was I wanted to test how the surfaces (glass and chrome parts) could be prepared to become „scanable“. As a relative cheap and easy to remove layer (only by water) I found chalk spray. For the complete backside of the car I needed one bottle with 500ml. This chalk spray was so finely ground that it even could be used to prepare a coin for scanning.

As you can see on the photo I used nearly the same set-up as for scanning the nameplate some time before. The camera was a modified Logitech Pro 9000 with a zoom lens:
http://www.david-laserscanner.com/forum ... 7805#p7805

And the motor scanner was shown here:
http://www.david-laserscanner.com/forum ... 9&start=30

The calibration corner was an a bit bigger one with a scale length of 99 mm and the tilted rod of the aluminium rig holding the scanner head was enlarged to 1m.

This time my neighbours came earlier, not just in the dark when they saw the rig and the laser line on the target but in the afternoon when I was spraying my car. I think some of them were worried about me.

Walter

Edit: The used calibration corner you can find here:
http://www.david-laserscanner.com/forum ... 8642#p8642
and the spray in the DAVID Shop:
https://ssl.david-vision-systems.de/sho ... h235s3cgq5

[Khalid Khattak]

Walter... I am really impressed by your great scans... Thanks for sharing and i always wish to see more and more scans from you...

[schmichael85]

Hi Walter,
I am in need of scanning a car roof, and I initially ruled out DAVID since I had no way of setting up the 90deg corner "underneath" the car, but from this thread I am convinced i can do this! I have been following your endeavours and they are truly impressive.

For your scan of the Corolla lettering, have you made an assessment or had a guess at the dimensional accuracy of your setup?

Do you think using a series of DSLR still frames i can achieve 1 mm to <1mm resolution? (i only need a roof profile scan in, say, 10 cm steps from the windscreen backwards)

[WalterMo]

Hi,
Sorry for the delay of answering. But I was on vacation and had no access to the Internet.
I think it's not so easy to scan a relative large car roof with a resolution of 1 mm or better. Try to scan it in one go. That means you should use the planeless scanning method. And for the reference scan you need a calibration corner which was made very precisely and has the same size as the roof.
I think the 90° corners of the garage were not perfectly stonewalled to use one of them.

Walter

[Khalid Khattak]

Walter.. we need some more scans from you

[bperez]

Walter,

Awesome set up! I am new to this, but I am particularly interested in obtaining full size vehicle scans as you are demonstrating in this thread. Please excuse my lack of understanding; however, I am confused about the process.

1) Prior to scanning, you calibrated your camera to the reference panel?

2) Given the relative postion of your camera and laser rig. I see that you can scan the entire backside of the vehicle. If you were attempting to scan the entire bumper the rig would have to be repositoned to scan the side. Once the rig is repositioned, is the camera recalibrated with respect to the reference panel? Are any reference makers required on the vehicle to assist in aligning the "back" and "side" scans?

Any details you can provide are greatly appreciated!

-BAS

[WalterMo]

Hi,
Yes, first of all it is necessary to calibrate the camera:
http://www.david-laserscanner.com/wiki/ ... alibration

Then a reference scan of the empty calibration corner must be made. Indoors or outdoors. After this you can even switch off your PC, the reference scan file is automatically stored.
But after this reference scan it is very important not to bump hard against the rig: The mechanical relation of camera and laser must be the very same! And it is forbidden to change the zoom or the focus adjustment of the camera.

Now you can carefully walk round the car with the rig and do lots of repetition scans without doing a new reference scan. But each time the movement of the laser must be the very same as during the reference scan. Means with the same uniform speed and starting from the same (angle) position of the laser's pan. For this job is a stepper motor (with gear) simply perfect. Markers on the chassis could be helpful later on to align the single scans, but they aren't necessary.

By the way, have you seen this thread of cnc Dave?:
viewtopic.php?f=2&t=1750&hilit

Walter

[bperez]

Walter - Many thanks for your input and reference to cnc Dave's thread. The explanation you provided clears things up for me.

BAS