View Full Version : Using a web cam for measuring

Crazy Ed
04-01-2007, 11:46 PM
Hey, guys, been away 'doing the ham radio thing' for awhile, and I mentioned using a digi camera to examine solder joints....especially on smt stuff. One of the ham's replied as a machinist and gave me this link from england to his program for hobby machining.


Looks like it might fit the bill for someone, maybe you, who knows.

Gotta catch up someday soon.

Weston Bye
04-02-2007, 08:12 AM
Some time ago we had a need to gauge a part in production that looked like a piece of bent coat hanger wire. No really good way of performing contact measurement. I came up with a fixture that positioned the part on a glass reticle etched with a scale with 0.1mm graduations. We used a video camera with a 50mm lens and extension tubes to view the portion of the part laying on the scale. everything was ridgidly fixtured. The field of view displayed on the 12" monitor was 10mm so considerable magnification was achieved. The image of the part on the monitor screen was masked with red transparent tape to indicate the high-low limits. This was a manual process, so subject to operator variation and interpretation. We eventually upgraded to an automated system that used the same principle.

04-02-2007, 10:14 AM
David Cofer came up with the idea last year. It is here:


I wonder if the British device is an outgrowth of that.

John Stevenson
04-02-2007, 08:29 PM
David Cofer came up with the idea last year. It is here:


I wonder if the British device is an outgrowth of that.

The original idea or at least one of them came from Tony Jeffree and Dick Stevens in the UK and was published in MEW in February 2006.

I built one of the first prototype ones with Tony and my files are dated October 2005 which fits in with the published article.

In January 2006, having been given details of the work done by Myself and Tony, Art Fenarty built this same idea into Mach3 and had a seperate video window which still exists today.
Mike later took this same idea and wrote his own software instead of the Willing Webcam software put forward by Tony and Dick.


04-02-2007, 08:46 PM
Noting this thread, one of the most accurate ways to make measurements of buildings and construction sites is digital photogrammetry but I know this would scale down quite nicely to parts. I did a bit of work for the National Guard testing this methodology for field survey work of training areas.

If you can take pictures from multiple angles with a calibrated off- the-shelf camera and feed them into some software that can correlate them, you can get a very accurate measurement. The software I used in my research is available from http://www.photomodeler.com/ Basically, if you can mark on each of a series of photos which points correspond to identical locations, the software can figure out in three dimensions the locations of everything visible to a high degree of accuracy which can be enhanced by taking more pictures. It does help to put either coded targets or very accurate round spots in the background of the picture around the item to be measured so that the software can more easily calculate the camera angles and locations requiring less work matching points between the pictures.

The photomodeler folks sell a package with a pre-calibrated Canon or Nikon SLR for under $3000. I have no affiliation with the photomodeler folks I just enjoyed using the software when I was doing that stuff.

02-05-2008, 09:43 PM
David Cofer came up with the idea last year.

Hmm. I've been using the Intel QX3+ children's microscope as a poor man's optical comparator for a long time ;)
I think several other folks here have mentioned using the QX3+ as well...

02-05-2008, 10:00 PM
Actually.. my software was developed on a Vision board in a At computer, you know 286 processor in 1988. THE input board cost $1800 or thereabouts then. HIGH grade stuff.. NOW it's a joke.

I was positioning a robotic arm using triangulation and orange super market dots. It'd search the screen and find the color from two cameras. Measure the cone of yarn and make a determination if it needed replacement.

During the move back from Colorado I lost key elements of the robot project. I also lost my $6500 a week income as a tattoo artist.

I didn't invent the center finder, nor the robotic vision, it was done by people a lot more intelligent than I. I did use it in my applications back in 1988 thou.

Now measuring, I found out you have fisheye in a lens, it does not linear measure across the image.. it wraps on the edges.

The same crap has been rewrote in five different languages now. AND, I have been studing C++ as one that started when I did, and still exists.

Now little of that matters thou, I had a nice 2 hour ride on the Burple colored Roadking today since it was 72 degrees here in Georgia.
Enjoy life, chase your dreams in your side time. I worked so much in my early life I feel I wasted it. DO THE THINGS YOU ENJOY NOW while your health is good.

02-05-2008, 10:21 PM
Ah, you're talking about the camera input and measuring software David.

I thought John and JC were arguing about who came up with the idea of using a web cam in the shop :)

02-06-2008, 09:29 PM
The last big thing I used the camera for? setting angles for drill bits. I had some templates I could pop onto the screen.

I was attempting to drill some 14" DEEP holes and they kept wandering off. I was thinking I could figure out why.

I hunted a loupe first with the angles in it. I had used a friends years ago. Someone here finally found me one. I didn't purchase it since the camera had worked so well.

NOW, a tool sharpener with a camera mounted on it.. that looks like the ticket. ALL the mills and drills you could see well what you are doing. I've been working on that for about three years now too tho, steppers are in a drawer.

I hardly have room to turn around in my shop.. Like I need another tool in the way? (yes I want one tho)