OK, following on from Durachko's thread I've decided to build a 3D laser scanner. I'm going to start by doing a turntable version similar to this one, but I want to work towards making a scanning head for my micRo. This will hopefully enable me to scan small parts for reproduction. The more I can work on it and develop it then maybe its capacity can be increased to cater for larger items, maybe with additional cameras/lasers.

The initial prototype will have a stepper driven turntable, cheap laser line generator, and a 2.0 megapixel USB webcam. I intend to write the software myself and to offer it up free of charge to anyone who wants to put together their own system. The basic steps will be:

• Record AVI file
• Analyze AVI file
• Output point cloud

If all goes well I will enhance the program to control the turntable and webcam, to fully automate the process up to and including generating an output file of the point cloud.

I'm only planning on going as far as the point cloud, as I'm hoping that there are programs out there that will take the data and convert it into one of the myriad of available 3D formats, be it NURBS, polymesh, or whatever. I've seen a couple that I will try, but first I need a point cloud!

I'll make it as accurate as I can, even if it takes 3 days to analyze a file! From there I can tweak it to find a reasonable trade off between accuracy and practical use. I've already thought of plenty of ways to drastically improve the software available from the above link.

Of course questions, comments, suggestions, and even criticism, are all welcome.

The web cam idea is OK but the sensors in those things are so full noise that you final product will have that noise in it.

To get the whole thing working the web cam may be cost effective but in the long run I would look for a used DV cam with a 1/3" CCD.

Sometimes you can find one with a bad tape transport and use a capture card to grab the frames. I have a good used capture card you could have free for testing.

A spinning laser might be sharper and brighter. A brighter laser will make less noise on for the camera sensor.

Neat idea I hope you can get it to work.

Appreciate the comments! But as you realized, the quality of components is just to get a prototype working. Once I get it up and running and start tweaking it I'll get a better understanding of the inherent problems and look at better quality components. It is DIY though, so I want to keep it cost effective.

I'll be working on removing the noise through the software, not quite sure how yet, but I have a few ideas.

I might look into finding a useable DV camcorder though. I might be able to control it straight from the software using Firewire.

You can eliminate a lot of noise by using a filter on the camera. A good dichroic one matched to the laser would be best. For testing though, just use a gel type.

DJ

Keep the software open-source!

Another friend just got a few line-collimated lasers to try the DAVID thing, this should be fun!

Some random inexpensive ideas from someone who knows absolutly nothing about 3D scanning.

1. If all your interested in is light from the line laser, would shooting the object in the dark help.
2. You might be able to eliminate a lot of noise from the image by raising the contrast. Both by software and lighting.
3. The colour and reflectivity of the object might make a difference (Paint it?).

----------------------------------------------------------------------------------------------------------
Those who can do. The rest of us are just arm chair inventors living off the glory of others.

The answer is mostly yes to all. Poke around the David link I posted. Interesting things in the forums...

Ditto.

I have briefly looked at the DAVID product, but not ventured into the forums yet. I'm way to busy with other stuff right now, probably for the next week or so. Once I have the parts ready I'm going to build a simple prototype, and then attack each of the problems one at a time. That's when I'll dive into the research. That may be bass ackwards, but that's my perogative.

The colour of the object is one that I've thought of, (and it's reflectivity, but mainly it's colour), and how that would react to the colour of the laser. One video I've seen has some excellent high contrast scans, and some extremely poor, very dim scans, all in the same capture file. So this is definitely an area that needs attention.

Once I have all the areas of concern nailed down, I'll look into better lasers, image capture, lenses/filters, etc.

So I checked out the DAVID site some more, and I'm actually glad to say that there wasn't anything there that I hadn't already thought of. I might not yet have a solution, but the fact that I've already been thinking independently of the issues gives me encouragement that I'm heading in the right direction!

Good to know. Be sure to send the software to me first.

No problem. Let me get it into Alpha first.

My webcam arrived today but not had chance to hook it up. If I'm not too tired tonight I'll try and write some code to control it. Should be 'interesting' as I've never written any C++ before, but how hard can it be, right? I'm sure a beer or two will help.

Got notification that my laser shipped today, so that should be here by the weekend.

Is it done yet?

I can see my pretty face on the screen. Does that count as progress?

Well I received my laser, but they substituted a different one. I specifically ordered a 5V one so I could run it direct from USB power, but they sent a 3.2V one. So what should I do? I think I have a 3.2V PS, but it might be 3.6V, and I don't really want to use it just to power the laser. What's the best way to step down from 5V to 3.2V? Just use a voltage regulator IC? (I know the 7805 is a 5V regulator, so I'm guessing I can get something similar that does 3.2V, right?) Well, off to do some more research. :sigh:

Resistor. It'll probably run fine on USB though after you account for the voltage drop...more important, how much did it cost?

Many of these devices are designed to run off batteries which can be (+) or (-) a volt. A long USB cable might also help drop the voltage within a tolerable range. alternatively if you don't have a resistor you could always patch a low voltage led into the circuit and add some bling.

I was thinking resistor, don't know why I didn't put that down as an option. It was late and had been a long day though.

It cost $7.50 and has a line generator lens included. Since it's 50% higher cost than your suggestion on the other thread I'm thinking it must be of the highest quality.


I'll dig out a cable and whatever resistors I might have and see what I can come up with.

Just run it on 5v!

BWAK BWAK!

Already done.

I haven't got anything done on the project though much to my chagrin. I live in Alaska, as noted, and, well, it's been an interesting few days!

Why is that your wife or something??

http://en.wikipedia.org/wiki/Sarah_Palin

Nope, I actually love my wife.

That was a joke I guess I was just trying to find out why in Alaska it was an interesting few days.

Oh, I realized it was a joke, not offended at all. And you're right, that's exactly why it's been an interesting few days here. I've been glued to my computer following all the crazy blogs out there. Got the square root of f!@k all done for the past week.

PAH!!! I've got the cube root of sweet f!@k all done for the past THREE YEARS!!! Amateur!!!!

Us mere mortals can only dream of becoming as completely unproductive as the mighty Durachko.