The recent posting which mentioned the toy nailbed, as used to give a 3D metal version of your face(ugh) got me thinking yet again about methods of extracting z axis information.

There are various diy(almost) ideas with links already in the forum, and no doubt they will get posted here also.
While waiting for some of them to come to fruition, I thought I might as well start a thread for other crazies like myself for whom ideas are easier to come by than cash.
This is not to put off contributions by eminent engineers, etc, but just be warned that extra points are awarded for the intellectual distance between one idea and the next.

Starting with the kid's toy, there seem to be two basic problems in getting something useful out of it.
1. Low resolution, unless you can produce a version with piano wire, and 32dpi is ok for you.
2. By extending the whole area upwards, little is gained except a metallic or reflective surface.
This would at least give you a surface with a consistent property to be exploited by the scanning method.

If the toy was reduced to a single line in the manner of the tool used for copying wood moulding sections, this then allows the raising of each contour above the level of the original.
Sideways scanning of this would allow a fairly simple optical method of retrieving the shape of the silhouette against an illuminated background.

A method like this has the possibility, given the resolution problem, of speeding up the production of z axis information, but still requires software to convert the file to a usable form.
Some interpolation might be acceptable, of course, and if you run your machine in raster mode, the programme might be easy to write.(Beyond me, naturally, but I'm sure someone will tell me how.)

My off the wall idea is to paint the object black, place it in a shallow dish, then just cover it with diluted milk, and photo it.
This will give me in a single step a greyscale image of the whole area suitable for converting to a raster carving programme.

Your turn.

John

John,
Really like your idea of the milk for acquiring images. Here in the US one can purchase ORGANIC diluted milk as the real thing. It's sold as 1%, 2%, or 5% fat content.

My idea:
I think that if someone could build a STEREO digital camera with electronics that could calculate the difference in the different pixels from the two images and would give an output for the X, Y, and Z coordinates, that we would be in business; That is, if we could afford the camera.

Anyone care to take on the task?

Jerry

Hi Jerry,
How about moving the camera between shots, or does that introduce too many possibilities for inaccuracies to creep in ?

Or split the view with an optical setup (binoculars plus mirrors/prisms)into a standard digital ?

Organic milk - I drink nothing else unless it's got ethanol in it. :D

I'll actually try the milk idea tomorrow. For now it's nightcap time.
John

John,
With a stereo camera, (it would have 2 sets of lenses and 2 different digital sensors) one would electronically have a true 2 1/2D view of the subject from one side only, not true 3D as stated in previous post.

Just think of the old stereo photographs of WW1 or WW2 days that the military used for intelligence gathering, and the viewer that was used.

Only a single photo would have all information needed to compute the topography for a 2 1/2D dimensioned part. The resolution could be near 1/1600th of an inch if the pixel layout was 1600 X 1600 which equates to 2.6 Megapixels. Many reasonably priced digital cameras attain 8 Megapixels now.

The camera would not have to output any photos, just process the info into a data or pointcloud file usable to create 2 1/2 D g-code.

I Have not given any thought yet to rotating an object and having overlapping photos taken. This action would require custom software which probably already exists in Military and NASA circles, but is probably not available to the general public.

Here is a website relating to a camera setup: http://www.ledametrix.com/
The crossed viewing is by crossing your eyes to see the true 3D images.

Jerry

http://www.camtronics-cnc.com/

Before you all re-invent the wheel....take a look at this website.

http://www.camtronics-cnc.com/

Before you all re-invent the wheel....take a look at this website.

Oh, I don't want to re-invent the wheel, I just want to go from a wooden wheel to a radial low profile tire with custom rims.

"Like doing jigsaw puzzles in the dark"

John,

Could you elaborate on these puzzling things that you are "working on" in the dark?

Jerry

Jerry, what are you doing up at 6 am or thereabouts ? Doing jigsaws in the dark ? No, I haven't tried it either, but it might be quite a challenge. :D

Well I tried the milk photo, examples attached. First the object - part of an old drawer handle. Then the photo immersed in 1/50 milk in water. Then a screen shot of the greyscale imported into Designer, the Carvewright software.
I didn't bother to spray it black, not that would have made a vast amount of difference. The crucial oversight, simple afterwords of course, is that the milk obscures detail as well as adding the desired to the greyscale.

I'd earlier thought of using a dye solution, and converting the depth of colour to a greyscale, but the milk seemed easier.
Anyway that's the next experiment, but will have to wait for a couple of days.

John,
When you get something that works for you, I want to know the secret since I need the same setup.

My son gave me a NEW FACTORY turnkey servo router system for christmas. He paid $3500.00 per the receipt. It is setup with 3 ea. Gecko 320's, 3 Pittmam servos with Pittman 500 cpi encoders, 8 TPI leadscrews, 2 120 volt switched AC sockets, and Everything just plugs togather, has limit switches, all cables etc. but some weird Supercam control software. I had it running under TCNC awhile ago, but it needs some setup and so forth.

He also gave me a new un-opened dremel tool and a new un-opened wet/dry vac that he had received in Christmas's past.

He is an outstanding human being. I taught him well, and it stuck. I am really blessed and also very thankful.

Jerry

John,
When you get something that works for you, I want to know the secret since I need the same setup.
....... some weird Supercam control software. I had it running under TCNC awhile ago, but it needs some setup and so forth........

Jerry

Well well. That's the same software that I'm in the process(lengthy as ever) of setting up.
I picked it up in the fall last year, so I suppose a year building isn't too bad by zone standards. The package was as per their kit - software, three steppers, a control box, cables etc., but no hardware as far as the machine goes.
The guy I bought it from threw in the computer that the software was loaded on as well for free. Like you say, it's designed to just be plugged together.
However the software is unusual compared with what most people use on the zone, with the g-code support seeming to be as limited as the designer thought necessary. Mine's the 98 version, and I see no season to lash out for the xp yet. Certainly a raster mode seems to be the way I'll start.

Back to 3d.
I'm next going to try immersing the same object, after I get some matt white paint today, into copper sulphate solution. This is the one simple innocuous solution that I can think of that doesn't dye the object(Even meths has a tendency to leave a purple stain).
I've just thought of using malt vinegar, but that would probably attack the metal.

It also occurred to me that if I put an object in a glass beaker of solution, resting on a turntable, I could then produce a complete 3D panorama of the surface. Having done the same with a plain cylinder as a reference, I could then remove the errors produced by the container.

Now the tricky bit, where we get to the "jigsaws in the dark" - what do I do with the images. I've no idea how to get from there to a 3D mesh/g-code, or whatever is required.
40 years ago I wanted to be a programmer, but never pursued it. There you go.
John

OK, I will bite, what about triangulation, three laser beams (what else???) very small resolution, software to move the beams and recognize when all three are all aligned (I never said this was gonna be simple), calculate the beam's relative angles and use trig functions to calculate depth... simple(chair)

Another method is to utilize a standard scanner, install a wing wong mounted to a doovalackey that will interpret the signals sent by the scanner head into a depth value. OK so the second idea isn't alltogether serious, but it is a concept nontheless.

Failing that you could use latex rubber, immerse the part to be copied face down into a pool of the stuff, wait for it to dry and use it as a mold.:stickpoke

OK... I got nothin'

Russell.

Hi Russell.
Your contribution is in keeping with my original intention of the thread.
Please keep us up to speed on your development of the doovalackey.
We had a wing wong, but it died.

I, too, wondered about using a scanner, but mine doesn't like working upsidedown, so I thought about converting the milk into a jelly, then I could turn it out onto the scanner glass. Not sure about the effect of lowering the lid, though.
Regards
John

I'm not sure if I'm understanding what you are trying to achieve, but it sounds fun.
Is this kind of thing any use?
http://www.brucerayne.com/scanz_explain.html

Regards Terry.

I'm not sure if I'm understanding what you are trying to achieve, but it sounds fun.
Is this kind of thing any use?
http://www.brucerayne.com/scanz_explain.html

Regards Terry.

Ha !!! I knew that lasers would be the answer, is there anything you cant do with a laser ?

Actually, this may not be that hard to do, well maybe not easy, but "possible".

OK, its added to the "to do" list, right up there with router # 5...

Russell.

Please keep us up to speed on your development of the doovalackey.
We had a wing wong, but it died.


You probably had a European wing wong, notoriously hard to keep alive, I was going for a Asia/Pacific breed, very hardy but nearly impossible to catch.

I, too, wondered about using a scanner, but mine doesn't like working upsidedown, so I thought about converting the milk into a jelly, then I could turn it out onto the scanner glass. Not sure about the effect of lowering the lid, though.
Regards
John

My biggest fear of that approach would have to be how to explain to my better half exactly WHAT I was doing if I got caught with jellied milk and the scanner.

In all seriousness, if you had a laser line, and took photo's with a camera, possibly from 2 or three or more angles, averaged them, recomposed the result, applied a smoothing mesh, the result might be usable. I'm thinking of a rotating axis with the piece to be "cloned" in the centre, fixed laser line to one side, and another rotating axis inline with the first with the camera mounted facing to the middle, with set or programmable number of angles/photo's per index position of the job. What would be handy is the parts needed to do this :

1 - Digital Camera : Show me someone who doesn't have one.

2 - CNC knowledge and/or keen interest : You are on CNCZone, you're hooked

3 - Stepper's : Lets see, no load, slow speed required, damn, floppy drive steppers would work.

4 - Laser line : You can buy them here for less than $60.00 and they come with a free circular saw.

5 - CNC router : OK so probably not needed, but I just happened to finish making one, and thinking of what project to build next...

6 - Software : This might be the sticking point, but surely there will be a "Guru" reading this that could help.

Might give this a bit more thought, but in the spirit of forward planning, I would think the machine needs to be large enough (in keeping with photocopiers and fax machines worldwide) to scan someone's butt... just a thought.

Russell.

I wait to see the first milk jelly butt.

Till then, here's my latest trial. Same piece but sprayed matt white(not very successfully - the top lines broke through the paint and are still dark), then immersed in mouthwash !
Flash photo, imported into Corel Photopaint, the contrast increased and converted to 8 bit greyscale. Exported to Designer as before.

I can see the possibilities now, and think the nextstep is to try a much higher concentration of colour than the mouthwash had. I'll probably try a spirit dye for the next test.

[QUOTE=epineh;226773] I'm thinking of a rotating axis with the piece to be "cloned" in the centre, fixed laser line to one side, and another rotating axis inline with the first with the camera mounted facing to the middle, with set or programmable number of angles/photo's per index position of the job. What would be handy is the parts needed to do this :

Russell,
I'm losing the plot here. (They all say I did years ago)
Fixed laser line - vertical ?
The second laser is giving me orientation and rotational axis problems.
Is the piece rotating ?
John

epineh-

That is exactly the idea I've been persuing over the last few weeks! I'm going to have some time this spring, and thought a nice DIY might be the design/development of a good 3D laser scanner... My concern has been in how accurate I could make the thing, after all we all want to be able to throw a hair in it and get a precise 3D model out, right?!?!

In all seriousness, I think using the multiple camera shot per orientation of the article and laser might help iron out errors... The machine itself wouldn't be too complicated, and the software to drive it not much worse (just steppers and tell the camera to take a picture, then analyze pixels for color of laser) but I haven't looked to far on the math behind the smoothing algorithms... Come to think of it, I'll start a thread on that subject... Be a nice partner with this one about probes, they'll go hand in hand!

John, yup the piece is in the middle of the rotating table, vertical laser line, with the camera set at say (just picked a number here) 30deg rotationally to the laser line. The camera will see a profile, compare that to a "calibration" shot with a laser line dead centre (might have to play around with a flat piece of cardboard or MDF or similiar), and you should be able to calculate a depth relative to centre.

As jabuffi mentioned, using varying camera angles will help to "sniff out" any dark spots caused by complex shapes, or simply to average out the profile for better accuracy.

Once the photo (or series) is taken of one profile, the piece is rotated by whatever resolution you think will be adequate, then so on...

If the depth from centre is worked out in software, you could use something like an Excel spreadsheet to do the averaging of the different camera angle values, then export this straight into g-code of sorts.

I guess the trickiest part will be converting the image data to a set of depth values, and I'm sure that could be done with little fuss. Just had a thought, you could include a "middle pointer" - a cone shape pointing down (centred) just above the piece being scanned, the laser line will show as an angled line above the piece, this could be used as a centre reference for the profiling software... I think you have stopped me from making my second router here John... ah well, I don't have enough room for what I had in mind yet any way :p

This will make an interesting Christmas project.

Russell.

epineh is right, an excel spreadsheet would be great for this sort of thing. One route to the depth values is through vector algebra... Establish the vectors that describe the plane of the laser and the separationof the camera as well as the many many vectors of each pixel of the camera's CCD, then scan the image for which pixel has the proper color (the color of your laser) With a little algebraic manipulation, one can extract what the coordinates of the point on the scanned object from these vectors (quick simultaneous matrix solution in VB) and go on and on!!! Limit to number of scans/points/degrees of rotation per scan will be camera megapixel size, storage size on your computer for point data (and if you want to keep the pictures etc...) and the time you are willing to alot to a scan.

The trick is making the scanner accurate, or at least precise and then applying the proper "smoothing algorithims" to take out/minimize any errors....

Now I'm thinking what would be better, a rotating table setup or a moving gantry setup like a scanner, for the piece you are doing at the moment John, I think a gantry would produce better results. Hmmm looks like I gotta build two scanners now (chair)

Ha jabuffi, got your post while typing this, I agree the depth would be a pretty simple algebra calculation. If the angles between camera and laser is known and constant, it would be a simple trig function to calculate depth. Finally, we found a use for mathematics hehe.

I am going to make an enclosure painted matte black inside to house all of this, might help sorting out inteference from outside light and reflections.

So now the question... where the hell do we post a build log for this sucker ???

Here is another thought, it would be pretty easy to retrofit an existing router to carry a laser and camera, keep the feed rates down nice and slow to stop camera shake, I gotta stop now, my head is hurting...

Russell.

To answer my own stupid question, I guess we post a build log in the same category this thread is posted...:wee:

So I come down to breakfast, log in only to find I got to start learning 3D algebra and how to use a spread sheet. I can feel one of my headaches coming on.
Saved by SWMBO also appearing for breakfast.

OK, I have a 50 year old memory of vectors, and I know what a matrix is, sort of, and that I casn use a spreadsheet to do the actual calculations oflots of data for me, but you're going to need to put it in pretty simple terms for me to keep up.

On the other hand, I still keep the dye-bath going as I can see that that could also be switched to the vertical mode and produce "in the round" scans.
Would it be possible to take say 6 dye-bath pics at 60degrees stepping, and "render " them onto the surface of a hexagonal column ?

So I come down to breakfast, log in only to find I got to start learning 3D algebra and how to use a spread sheet. I can feel one of my headaches coming on.
Saved by SWMBO also appearing for breakfast.


Hey... you started all of this, I was happy to start planning my 2nd router and possibly some projects for my first one, now I am looking around for material and spare steppers/drivers and planning layouts/concepts and software theories... oh yeah not to mention a 4th axis for my router to re-create the objects I intend to scan.

OK, I have a 50 year old memory of vectors, and I know what a matrix is, sort of, and that I casn use a spreadsheet to do the actual calculations oflots of data for me, but you're going to need to put it in pretty simple terms for me to keep up.


It has to be simple, or I'm out... I am only a dumb electrician remember...



On the other hand, I still keep the dye-bath going as I can see that that could also be switched to the vertical mode and produce "in the round" scans.
Would it be possible to take say 6 dye-bath pics at 60degrees stepping, and "render " them onto the surface of a hexagonal column ?

I'm not sure I get what you mean, but I think so, you want to copy something already hexagonal ? Mad if you don't. I am thinking of much higher resolution though, probably index half a degree at a time and have a vertical resolution of about 600 lines, it would be nice if my camera supported direct downloading to PC, sure would save time, although at 8 megapixel we are gonna burn some hard disk space... more power Igor!!!

Russell.

I 'm hoping to try the hexagonal idea because the dye bath produces the xyz information without the need to scan. By taking the six shots the camera will see down into the depths of any pockets.
Sorry, I've switched from bas relief to "in the round" without drawing a breath.

The initial idea for the immersion method came from the desire to use a piece of software(Designer) that can import a greyscale jpeg and convert it to a bas-relief carving done using a raster scan type cut, with z proportional to %black.(I don't have the machine, just solving problems for others)

I can see, though, that if I can get the detail picked up, the principle might also work for cnc lathes if I could produce a surface that could be wrapped round a "primitive" - in this case a hexagon if I took six pics.

Plastic figure about 4" tall immersed in black dye bath. I got a lot of dust on the surface of the liquid, so I've done a minimal editing on the image, then converted as before.
I think it shows progress, but still some way to go.
John

John,

Your projects are looking good. What kind of resolution do you think are you getting?

Jerry

i didnt read every single post here so sorry if this is already been mentioned or tried, but what if instead of milk, you used smoke or perhaps vapor from dry ice? it wouldnt have the same diffracting effect as the milk, and would perhaps preserve the sharp outines of the object. another thought is maybe with either the milk or the smoke, make it much more diluted, then take the photo and copy and superimpose it over itself many times in photoshop or paintshop etc. you would maintain the sharpness and still get your gradient.
Its all talk, someone let me know if it works.

Ed

Hi Ed.
The fundamental problem with my original idea of using milk is that the fat globules that form the emulsion scatter the light(they might even fluoresce in the flash, but that's another story), and so add opacity.
The idea is to "add darkness" with depth, not opacity, so switching to a coloured liquid will do that.
The smoke/dry ice idea has the same limitation as milk, but with the added problem of non-linearity - it doesn't stay still, and uniform in density, so there's no correlation between depth and darkness.

Jerry - not sure about the resolution. I think "poor" is the nearest figure I can come up with.
I'm going to try and improve matters from two directions now. The first is the lighting setup, so that I can avoid using flash.
The second is to look at an old photo programme that I have somewhere that enabled you to add two images together, rather like Ed has just suggested, but a bit more advanced. It allows you to set up a boolean filter, so you can add only the darker areas for instance. I think Ed's idea is sound, but does the same as increasing the contrast in the normal way. This would have the effect(I tried it in the first set of images) of increasing the depth of the image for the carving software, but does nothing for the resolution.

Busy weekend ahead, so it may be a few days before the next instalment.
John

B*STADS! You've all stolen my 3D scanner idea (not the milk one, that's just odd).

OK, here's my 2c worth...

1. 2 brackets on your toolmount, one for a laser, one for digicam.
(you already have a cnc machine up/down/alongy thing)
(every tool should have at least a couple of mounts).

2. A hi-res video camera would be best. Digital cameras are very slow
(well my piece of poo is). And need odd software or yucky TWAIN.

3. Capture video to an avi. (no special software needed).
as the machine scans over the 'thing'.
Bonus points for rotary axis useful but not necessary.

4. Insert aforementioned avi into a program that fits a nurb to the laser line in each frame. So for each frame (angle) you have a smooth line you can chop into data points anywhere you want. Build up an array them then even my mum could write a program to triangulate that!
You could do it without a laser if you want, just look for silhouettes.

5. Eat jelly and go play with wing wong.

If you'd read more carefully, 10bulls, and been a bit more sensitive, you might have guessed that our wing wong died when we gave it too many jelly nurbs.

Very sorry to hear that. Did I read that it was a European wing wong? You must have fed it wong jelly. I get my nurb jelly flown in from the Isle of Wight. So don't use Wong jelly, buy Wight jelly and watch your wing wong flourish.

This is a very interesting thread.
I'm trying to get my wing wong airborne, but was told I got the wong wing.
I got the wight jelly, but need to know, Is this the wong wing for a wing wong?"

Cheers.

No Terry, that's the Asian wing wong wing.
The European has a ring on.

John

The European has a ring on.
Ah! So that's what went wrong with my winged wing wong!

OK, I'm off to put on my thong and go play ping pong with king kong then off for a sing song at the King John son.

(...now where's my bong gone?)

When I first read this thread I thought senility had finally overtaken you greybeard. But who am I to judge! So in the spirit of experimentation...

Photo'd green man in a tray with blue food colouring (I think my hands may be stained for life!). I photochopped it a bit and made background blue.

OK, then I made a copy using only the RED channel. (If you've ever read 'The Deep', there's a good bit about the colour of blood at various ocean depths).

This last image was then stuffed into the CamBam heightmap plugin and hey presto.

Greybeard, you are quite possibly a genius!
Now all I have to do is persuade the kids to stay still under the bath water long enough while I go get the blue food colouring.

Looks like we need to go trade in all 3D modelling software for some trays, milk and assorted food colouring...

Hey John, are you the same Greybeard that has Blender tutorial videos on the Blender homepage? ... seems a strange coincidence.

Russell.

Not me, unless there's something about me that even I don't know yet :confused:

I downloaded blender two years ago, but haven't unzipped it yet - perhaps it's time I did !

John

Ah well it was worth a try, and I was hoping to get some free lessons !!!

It is actually a fairly capable program (and free!!!), though I am not at all any kind of expert on 3D modelling, it seems reasonably useful to me, I'm sure I read somewhere it exports STL files if needed, I'm thinking of using it instead of AutoCad, since I have to learn from scratch anyway.

Currently downloading what tutorials I can find, I think it could be used to incorporate laser line profiles from a series of photo's / movie frames... I just have to figure out how to make SOMETHING first...

Russell.

Hi Russell. I too have started on it. I've tried a few others before, but realise that some serious study is now necessary.

John

I've had another idea quietly brewing in the background which may appeal to those to wish to avoid total immersion in asses milk, or whatever.
The set up is extremely simple (always a dangerous sign) and I make no attempt to offer the software, but I would be interested to hear of any views on its feasibility, details of constructional improvements, and any offers of programming if it does seem worth pursuing.

The device is a laser "probe" that would be attached to a cnc spindle, for example, as a way of providing the x and y data.
Its construction (see diagram) consists of a laser pointer which can be "rocked" back and forth some 1.5 degrees to produce a short line of light on the surface of the target, some 10" below.
Looking vertically down through a pair of aligned pinholes onto the target is a photo receptor of as yet undefined quality.

The method of rocking the laser pointer also outputs a signal which is proportional to the movement, such as a pulse train.

When the laser spot is visible to the receptor, a count of the pulses from the zero position of each "scan" gives a number proportional to how far across the spot moved from its starting position. But within the limits of this simple setup, with a very small angle of scan, this will be proportional to the increase/decrease in height of the target at that point, each new measurement relating to the previous one.

John

I've had another idea quietly brewing in the background
I don't think this will work John. You are wobbling the laser in one plane and are assuming that light will reach the pinhole detector (located in the same plane) at some point during the lasers cycle.

Problem is, the scanning surface is just as likely to reflect light either side of that plane (i.e. infront of and behind the plane of your diagram).
You'd just get the odd pulse when the laser happens to be reflected in the same plane as the detector.

No, I'm afraid if you are going to be seduced by the sparkle of the laser, you must also accept that a camera of some sort will also be needed.

Good point.
I had assumed that whatever the angle the surface was, the photodetector would "see" the spot made by the scanning beam when it was below the pinholes.
I'm leaving out any considerations of resolution at this stage.

I did leave out a detail in last night's posting that the beam's zero position for each scan is the base plane of the target, directly below the pinholes. The maximum scan position would be to a point that would represent a 2" height, hence the 1.5 degree limit of its travel.

Perhaps I'll try it, if I can find a detector, or perhaps soomeone else could, if they have the kit ?

EDIT
Having had my breakfast coffee, sense has prevailed. If the resolutioin was 1mm, and it stepped once per second, it would take about 3 hours to scan an area 4" x 4".
Back to the asses milk, and improve my depth of field setup.

Good point.
I had assumed that whatever the angle the surface was, the photodetector would "see" the spot made by the scanning beam when it was below the pinholes.
Thinking about it again this morning, you may be right. I guess unless the surface is smooth/shiny you'll get some diffused light reaching the detector
(I've read coating objects with powder helps)
I wonder how laser ranger's work? (Don't make me google this early in the day!)

Re: Asses milk:
I quite like this method, but there are a number of drawbacks.
. The object needs to be uniformly the same colour.
. The object needs to be uniformly lit (without getting specular reflections on the surface of the liquid).

I have a bottle of blue lamp oil. It's for lamps. And its blue. Oh yes, It's made of oil.

If I paint an object with water based paint, then put it in the lamp oil hopefully the paint won't dissolve in the oil. This may increase the difficulty of making 3D scans of the kids though...

Step 1 (paint kids with white poster paint) :wee:
Step 2 (into the bath of blue oil) :drowning:
Step 3 (trying to avoid this...) (flame2)

Last thought before I go out to chop wood etc. - It seems like a compromise between the range finder and an ordinary scanner. Somehow get the scaner to use a laser line insteadof the light bulb, and change the scanning mirror to look down at an angle instead of straight down.
Ho hum, off to work.

It would be nice to "hack" an existing scanner to do this kind of work, but I can't imagine where to start. If you could seperate the data from each individual line as it scans, with the laser at an angle to create a profile. Looks like some serious googling is in order... and I was going to have an early night.

Maybe just make the scanner tray waterproof to hold milk lol

Russell.

Maybe just make the scanner tray waterproof to hold milk lol

Russell.

That's it Russell, why didn't I think of that.

Just put an aquarium on top of the scanner, put the target face down and cover with your favourite brew.
Mine's a pint of Nelson's Revenge :cheers:


John

That's it Russell, why didn't I think of that.

Just put an aquarium on top of the scanner


John

Mr Greybeard,

I think you will find putting the aquarium on top of the scanner a bad idea.

1, It will squash your scanner.

2 The moving light will stress your fish to deaf.

3 Fish don’t like Nelson's Revenge.

Hope this helps,

John

Mr Greybeard,

I think you will find putting the aquarium on top of the scanner a bad idea.

1, It will squash your scanner.

2 The moving light will stress your fish to deaf.

3 Fish don’t like Nelson's Revenge.

Hope this helps,

John

1. Use a small aquarium. Don't fill it up - just use enough liquid to to the job.

2. Give the fish some sunglasses, or wait awhile before scanning. See #3.

3. I don't believe that positive proof has been established in this thread that fish do not like Nelson's Revenge. Also, with enough alcohol and time, the light will not affect the fish -------- guaranteed.

Jerry

3. I don't believe that positive proof has been established in this thread that fish do not like Nelson's Revenge. Also, with enough alcohol and time, the light will not affect the fish -------- guaranteed.

Jerry

For the record, I and a number of my associated have surveyed over 17,000 fish and it is VERY clear from the results that not one fish liked Nelson's Revenge.

Regards,

John

I am surprised nobody has thought of a brain scanner (wedge) (chair)

Wonderful!

I got a good laugh and will try oit!

Hank

I am surprised nobody has thought of a brain scanner

10 points for the mental leap, but nil for not realising that very few brain scanners come with usb connectors. (chair)

I've just realised a major technical problem with my choice of Nelson's Revenge - gas bubbles. You'll have to find a brew that goes flat very quickly, unless you could remove the bubbles with a straw, of course. :tired:

Open to suggestions to solve this one.

you don,t use the USB port, just use the modem and post the code directly to this forum so that all can use it.

I've just realised a major technical problem with my choice of Nelson's Revenge - gas bubbles. You'll have to find a brew that goes flat very quickly, unless you could remove the bubbles with a straw, of course. :tired:


I thought you guys drank your beer luke warm any way, so it will probably be flat already :stickpoke

Russell.

I had a brain scanner but it could never find anything...:wee:

I think it was broken.

Russell.

How about a chisel and Eyeball MKI.

Only joking I used photomodelerPro5 to survey a plane in Helsinki the Heinoen HK-1. Attached is a WIP accuracy approx +- 0.8mm(when I have good angular seperation between shots) in all 3 axis's. Its very slow when doing this sort of detailed work. All points in this project were manually picked in each photo. Over 200 photos sometimes a few hundred in each. 3000 points so far not forgetting most of these are cross referenced several times!!!!!

Liam

I expect the festive season will provide copious quantities of flat brew, so I'll nip that debate in the Budwieser.

As I've now missed the last post for overseas, I'll wish you all a Merry Christmas and a Peaceful New Year from us, and all the Voles.

:cheers:

John

bigz1 - and what do you do in your spare time ? :eek:

john

John my better half would also like to know when I can devote some spare time to her. Which I reply she's too much like hard work.

Merry Christmas & A Happy New Year to everyone.

Liam

This maybe of use?

http://www.cs.tu-bs.de/rob/david.html

Looks really good. Will have to check it out in more detail after the Holidays.

:banana: Thanks for posting that link!
A new toy too play with!

and I already have all the parts...

Dave

Just went back and read the entire thread, Informative and quite funny
Will have to try that stuff submerged in tidy bowl trick

Merry Christmas to all

Dave

Allright, while I was going to launch into 2 simultaneous builds of 3d scanners, one gantry type, by using my existing router, and a rotary axis type table, I've conceded that I cannot achieve all of this at the moment, due to outside influences, mainly pretty much having no fixed address for the time being (Google cyclone Larry to get the idea), I am going to incorporate scanning into my 2nd router build.

I am thinking along the lines of a basic 4 axis machine, XYZ and a rotary axis for lathe type work, and possibly another 2 axis's on a totally seperate gantry purely for scanning. Sounds a little out there but thinking about it, the frame will already be done, rails done, controllers, power supplies and so on... if I make the thing longer by the amount it takes to fit another gantry, it's done!!!

A couple of more driver cards, and the rest is over to software. Imagine having a part on one end of a machine, being scanned, and at the other end a "clone" piece is being created real time in sync with the scanning process. May not happen quite like that but thats the dream.

The scanning gantry could be very lightly built, no cutting forces, little weight to shift, could probably be not much more that about 200-300mm wide.

Until I get the house (let's face it, the SHED!!!) built, I can't really do any building, but it leaves me about 4 month's to plan, and start on the electronics, pwr supplies, anything that is quiet to make and small in size. I intend to incorporate my own microcontroller based servo drives, just have to design/build/test/program/test them.

Don't really have a need to scan anything, but rest assured I will think of something by the time it is built.

Russell.

Don't really have a need to scan anything, but rest assured I will think of something by the time it is built.

Russell.

Go for it. Just store the scan in digital format also while scanning & routing.

I just wanted to get in on this thread so i would remember to think about it more later. Here is something that may or may not be of use to any of you. Good luck and I hope you've all figured this out by the next time look at this thread.

http://www.dasdistribution.com/products/laser_micrometers/index.htm

p.s. I'll do what i can too.

Sweet, thanks for the link.

Russell.

I did some more thinking, and googling, and found this site.

http://www.acuitylaser.com/USA/products/index.shtml

My idea, probably already thought of, is a 4+ axis CNC mill, with one of those lasers instead of a dremel. Or, maybe a hemisphere-thingy (it's a technical term, depicted below) where there is a laser that always points to the same spot but the laser moves. I'm not much of a programmer, more of an idea man, but surely there's someone out there that could come up with some software that can plot points by the 3d angle and distance from origin. I'm not sure if this makes sense, hopefully the drawing will get my idea across. If I'm re-inventing the wheel, will someone show me who else has done it?

Forgive my drawing, don't have any cad software on this PC, used good ol' MSPaint

http://img218.imageshack.us/img218/4585/untitledvu5.png

Another problem I'm having is the accuracy of the laser measurements. In some places it looks like they're saying its accurate to the micron level, others they're saying withing 1/8". Anyone know which one it is, or how I'm mis-interpreting it?

Russell - just back from Christmas with the grandchildren, and read the postings. I sincerely hope the new year is good for you "post-larry", and the new house/shed goes to plan.
Don't forget the go faster stripes !
Regards
John

Cheers John, having no house is one thing, but no shed!!! AAARRGGHHH!!!

We are doing fine, another 6 months and it's life as normal again. Did you mean stripes on the new machine or the shed? Not too sure if I want a fast shed, but worth a thought...

Right now I am just planning my next build and thinking of incorporating a second gantry for a scanner (in keeping with this thread), going for a more rigid build this time, no blurry laser photo's for this machine.

Has someone done the software yet ?

Russell.

dunno about the sw that you are thinking of, but have you looked at that link that bigz1 posted - the "david" scanner. that looks like serious stuff, and might be perfect for a 4th axis type.
I keep thinking that all the variations for cnc lathes are going back towards the old copy lathe.
friend and I built one about 40 years ago, where the follower was a 10" wooden disc to match the 10" circular saw which we used as the cutter !
We were making rocking horse bodies about 36" long, sets of legs and the head all on this machine built from scaffolding poles and a couple of railway sleepers as a bed.
Great fun to watch a circular saw directly mounted on the end of a 1hp moter swing on the end of a 6' length of scaffolding.
The major problem was the tool marks from the saw.
It took as long to remove as it did to carve the thing by hand, so that bit of Heath Robinson machinery never got us anywhere.
John

Edit
PS go faster shed of course, what else ?

The software I mentioned is for us to use once we have jury-rigged a laser to a machine, and got an endless stream of meaningless random numbers to perfectly replicate the original piece, including DXF's, STL's and G-code. All done in about 15 seconds, on a 486 running DOS.

Am I being hopeful ?

Just thought someone would have done it by now, after all, this thread is into its 5th page...

Russell.

Have you been at the joy juice ?
Or have I lost two pages ?

Have you been at the joy juice ?
Or have I lost two pages ?

Well it is up to five pages on my browser on its current settings, ran out of joy juice :confused:

My idea, probably already thought of, is a 4+ axis CNC mill, with one of those lasers instead of a dremel....

Forgive my drawing, don't have any cad software on this PC, used good ol' MSPaint

http://img218.imageshack.us/img218/4585/untitledvu5.png

Another problem I'm having is the accuracy of the laser measurements. In some places it looks like they're saying its accurate to the micron level, others they're saying withing 1/8". Anyone know which one it is, or how I'm mis-interpreting it?

Toby et al,
The Acuity line of laser displacement sensors operate using laser triangulation ~ meaning that they have a laser spot & a CCD camera that needs to 'see' the spot to be able to return a distance value. They are rated by depth of field (working distance) and the larger the range, the less repeatable/accurate they are. The laser spot size also increases, which can influence just how small the difference in height each point can be. As far as triangulation technology goes, there is no 'perfect size' for everybody. You have to pick a range and hope that it will work for the majority of the scanning that you need to do. The Acuity line of displacement sensors with power supply, cables etc will run you a minimum of $4,000, plus YOU will have to figure out how to implement it into your existing CNC control system. It's not the easiest of things for most people to do...this means programming in C++ or VB and getting it to sync with the XY values of your machine.

The drawing you attached is a neat concept, however it will not work properly with triangulation lasers such as the Acuity, Keyence or even (now completely unaccessible to the public...25pc min order) LaserRanger displacement sensors. Aside from sensor orientation, many factors need to be considered...like scanning direction (scans will differ greatly between unidirectional and bidirectional raster scanning) since the laser CCD sees will be sheilded when going up a 'staircase' shape compared to coming down one. You also will get a lot of data drop out with any angle greater than about 65°...it will not see a 90° wall or holes with high accuracy.

I spent about 18 months testing just about every type of non-contact system out there under $40k...3D cameras like the Minolta & InSpeck, took trips to check out ScanTech and other integrated systems...blah, blah....Nothing really 'did it' for me. The NextEngine is a neat concept, but really doesn't get it done either. I spent a good amount of time using an industrial type laser sensor and wrote my own software to integrate into my control system...I spent about $15k (doing it on a shoestring) and now get data in the submicron range from 0 to 6"+. This covers the majority of things that I scan in my business. If I had to go out and buy the setup today, it would probably cost something to the tune of $60k (with CMM/CNC).

My advice for most of you who really have a NEED for digitizing is to just send it out. There are quite few of us on here who know the ins & outs of 3D digitizing since it is how we make a living. Things like the David are neat for experiementing and hobbiests, but they are just too coarse for a number of us. Many things don't need to be laser scanned and can be done with a touch probe & a good working knowledge of a good 3D sculpting program like ArtCAM...I did this for years until scan time & accuracy became more important to me.

Just be careful & know what you are getting into before you plunk down your hard earned cash in hopes that a NextEngine, Acuity or other setup will be the silver bullet for a few grand...They aren't...

BTW, while it is faster...it takes time to scan. What used to take me 4 days with a touch probe that required cleanup, now takes about 4 hours with ZERO cleanup or data anomolies.

Just my $.02

-Brady

Thanks for the insight. I didn't know if it was possible or even worth it. Any info on the touch probe? I've never heard of it, but sounds like something i'd be interested in.

Hi Brady.
Thanks for your input. It's nice to have professional insights keeping our diy feet on the ground, even if our heads are well above the cloud layer.

When I started the post it was, by its very title, aimed at getting people to come up with novel ideas for putting together systems that might use low cost hardware in crazy new ways.
Obviously we're all keeping half an eye on what other threads have links to, and the link to the David method is certainly IMHO the best that I've seen yet.
Each method inevitably has drawbacks if thought of as universal. None is, so we keep hammering at the wall, hoping it will crack before we become unconscious, and we manage to get a glimpse of what is right for us, before slithering down into oblivion.
If we manage to describe what we have seen, and pass it on to others before the blackness descends, so much the better.
Otherwise, it's the band-aid and aspirin, and back to the hammering. :tired:
Regards, and best wishes to all for 2007.
John

Greybeard,
Are you using a touch probe on your CNC now? While it may be a slow process, if you set it up right you can get some really good scans that way. If you have a working CNC you can cut most of the parts for the probe on your machine & depending on your control software, be up and digitizing for just a few bucks in parts. The 'indoor flyer' version is posted on here a few times as well as a few others...It's not rocket science...just a 5-way switch. Heck, you can even cover something in tin foil and use a metal rod as a stylus (have ground wire coming from foil to machine ground) and get really good results...

I think that those who are interested in laser digitizing owe it to themselves to try touch probing 1st, just to see what kind of data you can get with low technology...Then you'll have a benchmark to compare what you get from the laser. I am amazed at how excited people get over the David laser setup...while it is cheap to put together and fast scanning, a touch probe will blow it away in scan quality...

-Brady

Hi Brady.
No, I haven't finished my own cnc yet, and am still learning a great deal about all aspects of creating machine made copies of my own handiwork.

One thing I noticed when reading about probe scanning was that the majority of users complained about the time it took. It's obvious that given even a fairly simple piece of hardware, a high degree of accuracy can be obtained if care is taken with the construction. But no matter how accurate the probe is made, this has no effect whatever on the rate of data acquisition.
The criteria must be the time/accuracy ratio for your own particular needs, coupled, naturally with the size of your budget.

In my own case that means a resolution of 0.1mm in a 10cm cube would be great. If that was derived from a touch probe, I'd be looking for 1 million points per face ?
Whatever the numbers, to use a system that takes a line rather than a point seems to me to have a head start.
regards
John

When using a touch probe, resolution is primarily dependent upon how accurately your CNC can move per step. Some of the smaller mills, like a CNC Taig etc, typically are geared for high precision at the cost of move speed. My CNC router can cut @ 700 IPM & rapid @ 1800 IPM...but using a touch probe it will only go about 60 IPM while capturing data & this depends on how tall (Z height) the object is. While is is technically *possible* to digitize at the resolution you described, you have to ask the ultimate question: Do I NEED this resolution? How many 1cm cubes are you going to digitize, or more appropriately, (Part A) what is the typical size of the parts I will be digitizing within the working envelope of my CNC and (Part B) What size bit am I going to use to cut the part after I have digitized it?

If you are going to scale the part up a factor of (let's say) 10X, then step resolution becomes an important factor, and you will have to tighten up on the stepover when you do the scan to avoid seeing stepover lines in your scaled up version. Once you start messing with digitizing & cutting on your CNC you will quickly understand the law of diminishing return and find the sweet spot for step resolution for the types of things that you do. You will also quickly realize that you are not going to cut something 6" tall (Z) with a 1/32" end mill for a number of reasons...

If you are a hobbiest, the cheapest way to get into the 3D digitizing game is with a touch probe. You can let parts scan for hours and even days at a time while you sleep, go to work and watch TV...all the while getting good data with little investment. I setup a scanning machine for a branch of the govt that scans things 2 & 3 weeks at a time! If you (A) have to scan things quickly and (B) HAVE to own the equipment to do the scanning, you are going to invest tens of thousands of dollars for this technology...Believe me...I'm a Scotch tightwad & pretty good at researching things...there isn't anything out there that will get you speed, resolution & good data on a low budget...a touch probe is hard to beat!

-Brady

John,
Since my thread was originally only a question about a software package, and we seem to be taking it towards the hardware end, I thought maybe we should move in on you!
I seem to have forgot to bring any Nelson's (nasty stuff!) but I can probably scrounge around for some Tetley's.:)

If you want to simply slap a module on an existing gantry, fire it up, and deal with the point cloud that it creates, then Baumer make some excellent Rangefinder style sensors. Set a CNC gantry to run a scan sequence, have a second computer wired up to the module to receive the data string. At the ends of the scan area, have an obvious marked edge to show where the numbers have to be broken to show a new scan line, when the scan's done feed the numbers into a spreadsheet to generate XYZ co-ords.

Seen it done by a factory rep, no fancy custom software needed.
Don't have the $1500 to spare for the toy.

For a quick'n'dirty solution:
http://www.pages.drexel.edu/~twd25/webcam_laser_ranger.html

For more info then you probably ever wanted:
http://www.repairfaq.org/sam/laserlia.htm#liarfi0

If you are a member of the cad cam dro group on Yahoo, do a search of the archives, a few years ago I posted some info on a fellow who was writting the software to use a rangefinder as a high speed non contact probe, it was either Bruce Larkin, or the Bruce (how Monty Python-ish, "and this is Bruce from the biology dept...") from New Zealand (the guy who got slapped down for his "how to" construction articles on a cruise missile).

Of course when you have your scanned model, then what do you do with it?
Why, you have a look here:

http://staff.bath.ac.uk/ensab/replicator/

ATB

Iain

..but using a touch probe it will only go about 60 IPM while capturing data ........-Brady

Hi Brady,
This drastically alters my perception of probe scanners. I had assumed from what I'd read that they would be aquiring data at about 1 point per second.


....... While is is technically *possible* to digitize at the resolution you described, you have to ask the ultimate question: Do I NEED this resolution? return and find the sweet spot for step resolution for the types of things that you do. You will also quickly realize that you are not going to cut something 6" tall (Z) with a 1/32" end mill for a number of reasons...

-Brady

If you have the time and inclination, a look at my web pages will give you an idea of where I'm coming from. ( www.fanmaker.co.uk )


Iain - re links, some familiar, some not, so thanks. I've got sam's laser pages filed away - part of my ongoing restoration of a co2 laser, and the reprap project is also being watched with interest.

In my "retirement", aka "my next career move", all this techno progress is leading me I know not where, but the journey is fascinating.
I started out with ambitions to replicate by cnc what I have been doing laboriously (and hopelessly uneconomically) by hand for the last twenty or so years. Now I'm looking for some alternative products, so the possibilities are endless. For the moment just my re-education is the driving force, but I am aware of the trap of it becoming the end in itself.



Happy New Year
John

Nice work on the Fans John, I now feel boring being just a "Sparky", my resume is nowhere near as impressive as yours. I bet a CNC router could make some nice frames for those !

Russell.

Happy New Year Russell - loved the Oz fireworks.

Fan frames is where I started curiously enough, but the possibility of producing carved period frames is mouth watering.

Currently playing with the hardware for the "David" setup having thought of using a projected white light. The attached jpg is a 1mm or 1/32" if you will line projected 4' away.
Someone asked for a better than1/8" at 4' but I've lost the thread(oh God, it's finally happened - first pun of the year )

Standard glass 35mm slide in plastic mount, opened up. I then cut a piece of ally kitchen foil to fit and layed it in place on the inner face of the glass. Using my trusty scalpel with a brand new blade, made a single cut through the al onto the glass, and closed up the plastic slide mount.


Edit - Sorry Iain, just tracked down it was your posting re 1/8" laser line that set me off.

Next job is to mount the projector on a hinged base to scan the test piece.

John

Aaaaagh..........just realised that depth of field could be a limiting problem.
With my old Braun Paximat 1700 projector and its 85mm lens, the depth of "reasonable" focus at 48" seems to be about 3" - 4". OK for bas relief, not so good for "carving in the round".
Increasing the projection distance improves that, but then spreads the beam wider. Trade off then seems to limit the possibilities, unless a better lens would give more depth ?
Comments from photo experts please.
John

John,
Neat stuff. At 1st glance, it would seem that most of the fan components are pretty much 2D, and can be done from 2D line drawings, using a cutting laser or rotary tool CNC with a very small bit.

The actual stands/frames could be done using a Roland Picza touch probe scanner, which was sensitive enough for me to scan a raspberry and other soft things...Just set up the scan and it does the rest with great quality while you sleep. You will then have a 3D relief that you can scale up or down or machine to size using a CNC with rotary cutter. You will need good software to edit & manipulate ANY 3D scan that you do since almost ALL methods of digitizing have some noise to them. I use ArtCAM and highly recommend it for 3 axis CNC cleanup, editing & toolpathing. You can pick up a used Picza scanner for between $500 and $1500 depending on model and who is bidding against you on Ebay.

I think that the David setup is a neat little science experiment, but it is far off from being a useful or practical tool for what you need to do. While I take my hat off to those who have developed it, it's still got a long way to go. The RepRap is pretty much in the same boat at the moment...it will be interesting to see how that project develops in years to come. Neither project is ready for prime time or suitable for your needs at the moment. The RepRap would be impractical for your products as it essentially would spit out parts that were similar to 'hot glue'...I think your stuff needs to be wood to maintain authenticity.

So...not to leave you hanging, if you have a particular design that you would like to have digitized, I will laser scan it for you for free. This way, you'll have something practical to play with when you get your CNC running. Contact me off list if you are interested, no strings attached.

-Brady

Kind offer, Brady, and thanks. I'll keep it by me.

Happy New Year to all our readers, guests included.

Now let's start the year with some serious ideas.
While seeing in the New Year here in Norfolk, UK, I suddenly realised that if I put the target into a straight sided glass, and topped it up with a pint, balance the whole thing on a lazy-susan, one might produce a panorama series of brownscale jpegs to wrap round a suitable central core.

The details are a bit blurry, but I put that down a lack of sleep. :rolleyes:

John

anyone tried using one of the new DLP projectors to project a razor fine on the item?

[QUOTE=greybeard;236581]Happy New Year to all our readers, guests included.

Same to you John..

--snip-- that if I put the target into a straight sided glass, and topped it up with a pint, balance the whole thing on a lazy-susan, one might produce a panorama series of brownscale jpegs to wrap round a suitable central core.
--snip--

And this would give you 3d... how?
Not trying to be critical, just can not picture the end result. What comes to mind, is there only being a very small visible area of the object in the glass, if you were looking to "map and wrap", then check out Statman Designs:

http://members.rennlist.com/statmandesigns/

He lays out the design in a graphics program and uses a cam program to wrap it around a cylinder. The blank is mounted in a rotary table and the mill cuts as a 2.5D setup.

I just looked it up in the archives, and he uses Vector cad/cam for the work.
I have saved the two pertinant messages saying how it's done, so PM me if you want them.

Doesn't help much with scanning a 3D surface, but it might be another way to get the end you are looking for.

ATB

Iain

New Readers and Honoured Guests.

Please take full note of the title of this thread, read the first page of posts, and then proceed with caution.

This is cutting edge insanity we're talking here. Cash is often the quickest solution to many problems, but if you haven't got any, and can substitute imagination, please stretch your grey cells and join in.

John

I thought this thread was totally serious, and to add one more totally serious idea, a good 3D scanner, in view of ideas already discussed so far, must include milk and lasers, this much we know. But what if the milk is coloured pink, as we know red is the fastest colour (fact.) and pink is milk with a bit of red colour mixed in, this combined with a red laser, possibly in a straight sided glass on a lazy suzan, controlled via a cnc driver of sorts, if you wanted to get picky.

This should make for the best possible scanner.

I was going to suggest red milk but that would just be strange.

Russell.

P.S. found more joy juice :p

...
Now let's start the year with some serious ideas.
While seeing in the New Year here in Norfolk, UK, I suddenly realised that if I put the target into a straight sided glass, and topped it up with a pint, balance the whole thing on a lazy-susan, one might produce a panorama series of brownscale jpegs to wrap round a suitable central core.

The details are a bit blurry, but I put that down a lack of sleep. :rolleyes:
...
Ah! Good to see you're back on track with such a sensible suggestion. I feared we were getting a little off topic with lasers, probes and other such unlikely sounding devices.
So I've tried balancing my pint on Susan but it kept rolling off the more curvaceous parts. However she was very good at not moving around much.
I noticed in Photoshop there are R,G,B and A channel filters. The A obviously standing for Ale, so I will investigate that exhaustively over the coming days.

I still have a trilobite hot glued to my CNC machine from a previous probing test. It got me thinking about an experiment I'd like to try some time. I'd like to progressively shave say 0.01mm off the thing and take a photograph of it at each step, like a poor man's MRI scan. I know this has been done for fossils before and it wouldn't be exactly quick. And alas, it would be a once in a lifetime opportunity for the trilobite. But it would capture all it's nooks and crannies like no probe or laser ever could.

EDIT: BTW Happy New Year!

hi Fellas.
Russell -
I strongly advise against irradiating pink milk with the laser. It will almost certainly gel, and then start crawling out the glass. Or possibly eating it.
10bulls -
You'll have to grind and polish the surface of the aforementioned trilobite, so I guess your looking at a cnc diamond wheel.
Your friendly lapidarist might lend you one, but then they might want your trilobite.

Off to the January sales to see if I can find a cut price cnc in M&S.

John

So I've tried balancing my pint on Susan but it kept rolling off the more curvaceous parts. However she was very good at not moving around much.

There is a problem with that approach, after a while the beer gets warm, not such a problem in the UK, but in Australia is a criminal offence. Also beer CANNOT be wasted.

I still have a trilobite hot glued to my CNC machine from a previous probing test.

I think that kind of behaviour might be illegal in some countries


I'd like to progressively shave say 0.01mm off the thing and take a photograph of it at each step,

See I think shaving Trilobite's is kinda illegal as well, I am left with several disturbing mental images here.

But it would capture all it's nooks and crannies like no probe or laser ever could.


I think someone's been visited once too many times by aliens...

Russell.

I noticed in Photoshop there are R,G,B and A channel filters. The A obviously standing for Ale, so I will investigate that exhaustively over the coming days.



Well spotted, I will also investigate this phenomenon here - purely for scientific reasons, the whole northern/southern hemisphere thing, we must stay objective.

Russell.

New Readers and Honoured Guests.

Please take full note of the title of this thread, read the first page of posts, and then proceed with caution.

This is cutting edge insanity we're talking here. Cash is often the quickest solution to many problems, but if you haven't got any, and can substitute imagination, please stretch your grey cells and join in.

John

Teknikly called Brain Storming

I'd like to progressively shave say 0.01mm off the thing and take a photograph of it at each step,

See I think shaving Trilobite's is kinda illegal as well, I am left with several disturbing mental images here.

Russell.

This have been done with a frosen human body, they frose the poor bastard in some blue stuff and then started shaving off 0.1mm, and then took a photo.
Cant remember what the project was called.

Found it http://visiblehuman.epfl.ch/index.php (thanks Google)

Quote "I don't wish to know that. Kindly leave the stage" Unquote.
This wasn't what I had in mind (!!!) when I said "stretch your grey cells".
John

Gentlemen,

What would be worse:

(A) being the shaved specimen,
Or,
(B) being the one who has to clean the machine after?

Either way, it is a disturbing mental image...

John,
I was trying to stretch my grey cells and many of them broke; Some became frazzled and limp in the middle, and others took on a different neurotic shape.

Time will only tell what the end result will be.

By the way, what is the best type of alcohol to repair my damaged and distressed brain cells? I was thinking that a true natural corn based remedy might work really well. It works LIGHTNING fast I am told. One might really come up with some brilliant ideas if this cure works as well as the historical records indicate. I have heard that a complete idiot can become a complete genius in just a few hours of partaking of this corn spirit.

Anyone with experience in repairing damaged and distressed brain cells?

Jerry

Sounds like some reverse engineering is needed. Try Methilated spirit colored to your own taste with boot polish.

So we've moved on to a new idea then - how to use three glass containers filled with a coloured spirit, possibly any bubble in them causing a potential problem, arranged in some comfiguration that enables us to extract three lots of data to line up the cnc machine head in spatial terms. That's my level best description.
It'll never work.
John

You are right, it will never work, no mention of lasers...
Russell

To get away from the serious postings, and be a little silly for a moment, does anyone know of a piece of freeware that will enable me to "subtract" one greyscale image from another ?
If my original target has a` lot of dark colour on it, this will add to the "dark" value of the depth of the coloured solution.
If I were to also do a "dry" scan, I could take one from the other, to just leave the depth information.
I used to have a programme on my first pc running win95 that did just that, but I've never seen the same facility since. I think they referred to it as "Image calculation" or some such name.

John

Edit - it was called "Photostacker", written 1994 !

erm, I just opened photoshop, clicked on images then calculations, a small window opened, and had source 1 at the top, source 2 below, then a blending drop down box that had options like multiply, blah, blah and near the bottom, add and subtract !!!

man I'm good...

good to see some more light hearted posts back here ;p

Hmm just re-read you post, you wanted freeware, I'm kinda thinking photoshop is not really freeware...(chair)

Russell.

Hi Russell - thanks. I've got ps5 somewhere on floppies, so I'll see if it's on that(when I've found a floppy drive that I can use etc.)
John

Edit just found it on this laptop :wee: so i'll see what i can do.

Edit2 yes, there it is, so I'd better go and build the aquarium over the scanner.
Sudden brainwave - my elderberry wine is much more intense in colour than anything that came out of a pub, so I should get a greater colour change/depth ratio. I hope this will mean a better resolution.
John

Am I right in assuming that the wine is best used for this process ? I know I've had a few turn out like vinegar. If not so long as whatever you are scanning doesn't dissolve you could always have a drink from the leftovers while you are composing the picture on the computer.

Well, I made 15 gallons last year, so I reckon I can afford to experiment with some to see if the taste is affected by which target materials.
:D
Going into town soon to see if I can find a suitable plastic container, otherwise it will be a shallow glass bottomed tray. Just a wooden wall round the outside sealed with silicone bathroom caulk. Place on the scanner and see what we get.

John

Well I finally got round to doing the experiment.
First problem, my wine is a bit too dense in colour. It's virtually opaque at 1/2" deep.
I'm using a sloping white plastic ramp as a test piece, and it disappears in the wine too rapidly.

Second problem, I've now discovered a curious distortion by the scanner. The sides of the test tank appear to slope in at the sides. I think this is barrel distortion - fairly appropriate I guess. (chair)
On second thoughts perhaps it's "pin cushion", can't remember too clearly.

It means that only a narrow strip up the centre of the scanner is not distorted, so unless I can see how to correct this, the scanner is a non starter.
And I thought it would give a flatter image than the camera for larger objects (talking bas-reliefs here, not sculptures).

So it looks as though it's got to be the camera. Possibly a better focus, but a more elaborate set up - camera on floor looking up at the bottom of the target tank.
Mmmm ?

John

I'm not sure i understand exactly what you guys are doing with the liquid idea, but i had an idea I thought I'd toss out there anyway. Could you paint your object white and use motor oil floating on top of water?

Oil on 3d waters ? Mmmmmmmmm?
Could you fill in a few more details please Toby.

I'm afraid the only way to understand what's going on here is to read the whole thread, especially my first post #1.

When I started this, I tried to be very specific, but inevitably various contributors wandered off in all directions.

3d scanning is becoming more interesting for cnc work as well as being a bit of a buzz word generally. If you've got $$$$ (or any other currency) to throw at your current problems, great.
But for the many of us that don't, ingenuity is a common substitute, and ingenuity comes with practice.
Building on other peoples shoulders is perfectly acceptable, so please add your ideas in detail.

John

The way i understand this idea is to get a part (color a) and a liquid (color b) then put the part in a container with a scanner under it. Then, slowly add the liquid (enought to move the liquid level up a very small amount) while constantly taking scans. If the difference in the two colors is distinct enough, you should be able to take the successive outlines, stack them on top of each other, (software way out of my league) and have a 3d model of your part.

The problems I've seen are mostly due to the liquid being too opaque to get much detail at depths. But...

If you use a thin layer of oil (enough to black out light, or just scan in a dark room) on top of water, the opacity shouldn't be too much of a problem. Just keep adding water like you normally would. Or, slowly take water out so the oil doesnt discolor the part that has yet to be scanned...

I don't think I'll be getting around to this project anytime soon, too much on my plate already. I'm probably going to build a small engraver with enough z travel to attach a touchprobe and do my scanning that way. If you ever get this going, let me know.

Sorry Toby, you seemed to have missed the principle I'm using entirely.
The depth of the coloured liquid is directly proportional to the density of colour shown in a single scan.
Double the depth and you double the desity of colour.
Convert the scan to a greyscale, and numerous programmes can convert the grey density back to a depth(Z axis) distance.
I hope this makes it clear.

My immediate problems are two.
1. the wine is so dense in colour it is virtualy opaque(ie "black" after conversion) within <1/2".
This is solved by diluting the colour with water.
2.The scanner has a distortion, apparent only when you are looking at an image above the level of the glass top.
As yet no solution .

John

John,

Any luck with your little project?
I've been given a sample of the Baumer laser range finder to play with (after placing an order for approx $20K in electronics you would think they could have at least offered the software to go with the free sample!!), so everything else has been shelved lately......"David" was less frustrating to deal with!

ATB,

Iain

Hi Iain.
No, the big project has taken over for the time being i.e. building the 25ft long greenhouse before SWMBO needs to get the seedlings in !

Regards
John

As yet no solution .

John

Drink more wine !!! (or beer)

Russell.

Drink more wine !!! (or beer)

Russell.
After new year I thought I'd have a bit of a health binge. Laying off the drink and chocolate, eating my fruit like a good 'un. Even cutting out coffee!:eek:
I then proceeded to come down with the worst flu I've ever had. I even resorted to seeing the quack monger, er I mean doctor, it got so bad.
Hmmm maybe nature is trying to tell me something. :idea: Though I can get into my old jeans now. :D

Anyway, dragging this post kicking and screaming back on topic. I had a little brain wave :wave: on how to get CamBam to output points from the heightmap plugin.

First I generated a heightmap as per usual. This generates a line as shown in the 1st picture. Select this line if it is not already.

Now do Insert->Pointlist->Step Around Geometry.
This will insert a point along the line every N step distance.
By default, the heightmap will do 1 bitmap pixel = 1 drawing unit (This can be changed in the heightmap options).
I entered 1 for the step distance then pressed OK.

CamBam currently displays points using biggish squares so it will look cluttered, but don't worry about that.
The line object can now be deleted.

I then exported the drawing to a DXF file. The second picture shows this in Autocad.

The heightmap generator itself is still as dumb as 3 pints of elderberry wine but hopeful that will improve this year.

Hope this helps.:cheers:

Hi 10bulls, Russell, and all.
I'm still suffering the aftershock of my car getting through its MOT without any problems !(Annual road worthiness test)

Keep up the good work on remodeling the waistline.

When I can escape from the "crystal palace" construction, I have a mind to set up the camera underneath the wine cell, thus combining two ideas into one. Gets over the distortion of the scanner(doesn't appear to be a solution to that one), and with the manual control on the settings, I should be able to get the required depth of focus.

Regards
John

..I have a mind to set up the camera underneath the wine cell, ...
It's about time somebody kept an eye on those sneaky alcoholic (yet slightly posh) terrorists!

Having just poured myself out a sloe gin(good brew this year), and re-read your posting, I realise that there may be cause for concern with my ever closer association the the wine cell. (group)

I'm drawing the line at filling it with sloe gin, even though the colour is better for this application. :)

John

..I realise that there may be cause for concern with my ever closer association the the wine cell. (group)
I wouldn't be overly concerned. I heard on their last attempt to blow up a bus they burnt their lips on the exhaust.

Thanks. (Much laughter)

10bulls & Greybeard,

10bulls: Enjoyed and still laughing about the post relating to the exhaust pipe.
John: Same goes for the post on the Wine Cell.

I could really enjoy sitting down with you guys and having a few brews. You guys really made my day.

Jerry

Any more progress?

I have been busy on my little photomodeling project and hope to start doing the final drawings in Solidworks soon.

Hi Bigz1.
No, I'm afraid too many other projects are getting in the way of my pushing my own ideas forward.
I'm still waiting to try out a circular light/light tent mentioned somewhere in the thread. This will mean actualy tripping over a suitable fluorescent tube before I'm diverted from the current project - a 28' greenhouse for SWMBO.

Glad to hear of any new ideas that come to the surface though.
Regards
John

This will mean actualy tripping over a suitable fluorescent tube before I'm diverted from the current project - a 28' greenhouse for SWMBO.
John

Haven't you finished that greenhouse yet ? And I thought it was 25 ft long :confused:

Maybe you could post a build log here on the zone, so we can help you along the way, I guess it isn't really specific to CNC, you might just have to put a servo or stepper on an angular axis to open the door...

Russell.

Haven't you finished that greenhouse yet ? And I thought it was 25 ft long :confused:

Maybe you could post a build log here on the zone, so we can help you along the way, I guess it isn't really specific to CNC, you might just have to put a servo or stepper on an angular axis to open the door...

Russell.

This is my main problem. The da*n thing keeps changing size and shape.
Each time I measure it, or try to square it up, it seems to have moved.
Tomorrow I bolt the thing down to its foundations, so hopefully any discontinuity in hyperspace will also be fixed. (These are serious bolts I'm using)

John

Hello everyone,

Very interesting thread!... in googling around I found the following site which may be of interest as a potential starting point,etc.

http://www.muellerr.ch/engineering/laserscanner/default.htm


Brent

Hi Brent, and welcome.
I had a look at the current site that your link points to, and remember seeing some of his work last year. The current site seems difficult(to me) to get any info out of other than being a gallery of what he is doing, where last year it seemed to be more informative on the "how to".
Perhaps he's naturally protecting the possible commercial aspects of some of his work, but the regime here is strictly open source ideas.
The more "off-the-wall" your ideas, the more Brownie points you get (does that translate to US ?), double if you can add practical details, double again if it involves any kind of alcoholic beverage. :cheers:

Most of the contributors are up to their blood-shot eyeballs in other projects, so the thread goes in fits and starts, but with the Spring starting to warm up the grey cells, I live in hope of a breakthrough this year.

Regards
John

:stickpoke

Thought I'd give this thread a little poke in the eye with a stick (couldn't find any pickled eels :))

Has anybody ventured any further into the world of 3D scanning ?

I still have it on the list of to do things, but there are just soooooo many things in front it isn't funny. But there is light at the end of the tunnel... I have been playing around with Linux for my EMC based controller and am totally impressed with the OS.
So much so I am currently converting my home PC's to Ubuntu. One of the things I like is the ease of installing programs and the sheer amount there is to choose from, and there is a lot of programming language support, which getting back to this thread might give me a little incentive to start the software side of things, maybe.

Anyway, I'm off to see if I can cut the list down a bit, if I can get it under 1000 things to do I will be happy :)

Russell.

...
Anyway, I'm off to see if I can cut the list down a bit, if I can get it under 1000 things to do I will be happy :)

Only 8 things to do! What are you complaining about!

Hi Russell.(If your still up. If not "Good Morning tomorrow")
I had another brief look at the David site where they seemed to have upgraded since my last visit, and more importantly, have a stiching programme available at small cost. When my list gets down to 010 projects, I'll probably have a go.
I'm very interested to hear more about your ubuntu experience. I was going to post to you having picked it up on another thread, so I'll enquire here instead. It's something that occured to me to do when I was first faced with xp, but the carvewright machine that I'm interested in is only xp at the moment. That may change of course.
Anyway, now vista is looming and the "gates" seem to be closing on all degrees of freedom, I think the time has arrived to bite the bullet.

Regards to all.
John

Only 8 things to do! What are you complaining about!

Ha, I had to read that twice before it clicked lol, my excuse is it is late here, I am sticking with that :)

John, I have to say that Vista was the final straw, I can't say I have used it, but by the reccommeded hardware to run at the "same" speed it is already looking bloated.

I also have to say the laptop I have just installed Ubuntu on (it is SWMBO's - I live dangerously) is running about 4 times faster for general stuff (estimate only, but it was almost unusabe previously - new laptop, not high end but not that bad - 2 months old). It had XP Home installed from factory, and was a real PITA to use, it just took forever to do anything, even opening a browser was a slow old procedure.

Things are running fine now, I am using it to post this, but there was a bit of a learning curve, it pays to have a seperate machine on the 'net beside to get help for starters. I am using the live install version of Ubuntu 6.06 for EMC, but was so impressed with the ease of install/use after installing on my router PC, I decided to put it onto other machines.

The only thing that proved tricky was getting internet connected (of course) but once I worked it out, it is pretty simple to setup, main prob is that it is setup for ipv6, the new protocol they are supposed to switch to when they run out of IP addresses with the existing ipv4 protocol.

It will actually connect to the net while running in "live" mode, no HDD install at all, I thought that was pretty cool.

Good luck if you decide to have a go, my advice would be to get an old "clunker" PC, say a 2gig processor with about 256 meg of RAM, and whatever HDD you have lying around (more than 10gig) to experiment on, no doubt you will have a little "fun" before you sort it, but it is worth a try, I am glad I did.

Russell.

I love it! The milk and colored water idea was something we fooled around with a decade or so ago. Soy sauce also works for very shallow objects. These liquids worked well but we got tired of painting everything white or some other uniform color. That and the food coloring stained some models.

So we added a white latex membrane to keep the model dry. This worked but you would lose fine detail. However, a texture map captured from the original object could cover up a host of geometric errors when used as a model in a 3D environment.

Next we mixed up some elastic polymer instead of a liquid. We had pretty good luck with polyurethanes and various types of RTV rubber. It's tricky finding compatible dyes and degassing the polymer as it sets up.


For the laser scanner idea, has anyone tried mounting a Sharp distance sensor on an XY table? I would use a GP2D120 (http://www.acroname.com/robotics/parts/R146-GP2D120.html) analog output sensor.

_I just wanna ask: Does anybody use DeskArt (http://www.deskam.com/deskart.html) software?

I love it! The milk and colored water idea was something we fooled around with a decade or so ago. Soy sauce also works for very shallow objects. These liquids worked well but we got tired of painting everything white or some other uniform color. That and the food coloring stained some models.

So we added a white latex membrane to keep the model dry. This worked but you would lose fine detail. However, a texture map captured from the original object could cover up a host of geometric errors when used as a model in a 3D environment.

Next we mixed up some elastic polymer instead of a liquid. We had pretty good luck with polyurethanes and various types of RTV rubber. It's tricky finding compatible dyes and degassing the polymer as it sets up.


For the laser scanner idea, has anyone tried mounting a Sharp distance sensor on an XY table? I would use a GP2D120 (http://www.acroname.com/robotics/parts/R146-GP2D120.html) analog output sensor.

Hi Peter, and welcome to the club. I feel honoured that you've made your first post here.

10 extra points for the membrane idea as a sealer before immersion.

I'm hesitant about the idea of using a single point device like the GD2d120.
I've had a look at the spec and noted a change of about 0.2volts per cm in the middle of its range. If the response speed is high, I guess this could be an interesting way to go.

John

Hi Peter, and welcome to the club. I feel honoured that you've made your first post here.

10 extra points for the membrane idea as a sealer before immersion.

I'm hesitant about the idea of using a single point device like the GD2d120.
I've had a look at the spec and noted a change of about 0.2volts per cm in the middle of its range. If the response speed is high, I guess this could be an interesting way to go.

John
Thanks. I've been lurking for quite some time now. Homebrew 3D scanners are something I've tinkered with off and on for years.

The response of the GD2d120 is pretty slow. I think its somewhere around 10-50ms. It would be fast enough for an overnight job though.

The projected light line scan idea mentioned earlier in the thread was something else I worked on. We built a cylindrical line scanner like one of the Cyberware (http://www.cyberware.com/) models. The hardware was pretty simple: some sort of line generator, 2 video cameras (for redundant depth information), a rotating platform, and some custom software. We used a 486 computer to process the images and build up a 3D model. The actual capture platform was large enough to rotate a 200lb man and was built from random plumbing parts gathered at Home Depot.

We used a white light projector to capture a registered texture map at the same time as the object geometry. This was before inexpensive video projectors, so we used a slide projector with a pair of razor blades stuck in a slide frame. The slight gap between the blades created a very straight line of white light.

If you only have linear scans of sides of an object, multiple range images can be "zippered" together to create a complete 3D model. This can get pretty complex. Its been years since I looked into this, but there was an open source package for doing this at one time. Google "zipper range image".

The advent of cheap video projectors makes building a more complex structured light 3D scanner an attractive alternative. The projector can be used to cast multiple lines, a grid, or even sweep a line without external mechanics. There have also been several approaches that cast various checkerboard patterns. All these approaches require some sort of custom software, but it's not terribly difficult.

_I just wanna ask: Does anybody use DeskArt (http://www.deskam.com/deskart.html) software?

Yes, I have used DeskArt before but I had limited success with it. I wasn't very good at cleaning up the file and my machine would only cut at about 1" a minute so I couldn't be bothered waiting 4 hours for a 2" square cut.
There was a site on the web a few years ago run by a chap by the name of Marty Harper who had a lot of samples and some were of deskArt. They looked very clean and very similar to the original and I think the secret was in the dye or coloured wash that was applied after the cutting. He used Granite and Marble for his cuts with an old Vertical mill converted to CNC.
I haven't had time to get back to it for some time but would like to in future. It's a shame he doesn't have his site up and running now though.
DeskArt can do good work but it is all in how you clean up your file and add lighter bits and darker bits for the hight map. Marty had some samples of what to do but they are long gone. I may have saved them somewhere but it would take me a month of Sundays to find them in amongst 100 disks.
Richard.

I just came across this (http://www.instructables.com/id/E16Y5T8F22U9YWA/?ALLSTEPS) and thought of this thread. I don't have time to mess with it just yet but maybe someone else does.

Cheers.

I just came across this (http://www.instructables.com/id/E16Y5T8F22U9YWA/?ALLSTEPS) and thought of this thread. I don't have time to mess with it just yet but maybe someone else does.

Cheers.

Thanks for the link 50ohm.
Like you, time is at a premium at the moment, but a quick look found it intriguing.
If I guess right, he's using the milk to generate a series of contours which are then used to create the 3d image.

If this is so, the idea is one that falls neatly between "whole image" recording, like the total immersion in wine idea, and the spot scanning of the touch probe.
I suppose this, like the David et al methods of line scanning, would suggest there are three possible approaches - spot, line, or whole body methods.
Put more succinctly, 1, 2, or 3D input methods. Perhaps the line method should be split into straight line and contour (2 and 2.5D ?)
Anyway, you've got the grey cells going again.
Regards
John

Uh-Oh, looks like you have got John going again Fifty Ohm... now the price of milk in the UK is going to rise due to lack of supply, anyone over there should stock up now :D

Russell.

On the other hand, the elderflowers are just about ready to pick, so perhaps I'd better concentrate on my priorities. :cheers:

John