The thought of using camera's sounds interesting. Measuring angles with camera's normally would involve a delicate, costly and complicated method of skewing the cameras. But the same results may be possible with fixed cameras and using some form of image processing to determine the angle of a single target on the two cameras. Using a left cameras and a right camera, a fixed distance apart, and counting the numbers of pixels between where the target is on the left camera image and where target is on right camera will give you a relative angle measurement. As you know the distance between the two cameras, you can calculate the distance For height, count the number of pixels between the top and bottom on a image from one of the cameras. And using the calculated distance, determine the height. This concept could be tested by setting up two targets (repressing the two cameras and at the same distance apart that the two cameras will be mounted) and taking a picture with a single camera at difference distances from targets. Then counting the pixel separation between the two targets on the single image. This pixel difference could also be used for a table of pixels vs. distance. Bill ----- Original Message ----- From: "Rolf" To: "Microcontroller discussion list - Public." Sent: Thursday, March 02, 2006 6:27 AM Subject: Re: [EE]: Distance and height measurement. > What about the recent work done on the Shuttle, where there was a > "laser-camera" used to inspect the ceramic tiles once the shuttle was in > orbit. > > If I am not mistaken, the process used parallel laser beams to create a > reference on the tiles, and then the reflected light could be used to > read the reference beams. > > The observed distance between the reference marks would be inversely > proportional to the distance of the reflection from the observer. > > There was an article recently where it is now being used in crime > scenes.... > > Here...: > http://science.slashdot.org/article.pl?sid=06/03/01/1912255 > > Rolf > > Howard Winter wrote: >> Tony, >> >> On Tue, 28 Feb 2006 20:50:39 -0600, Tony Harris wrote: >> >> >>> Hi all, >>> >>> I'm tinkering with an idea, but I think it is beyond me - I'm not sure >>> tho. >>> Basically, it's a height at a given distance determination. >>> >>> I was thinking of using 2 sensors one angled down and one angled up, >>> using >>> the angle between them, and the distance calculated from the sensors to >>> determine the height. >>> >> >> Basic triangulation - measure the horizontal distance, and the angle from >> horizontal to the top. I did it >> once to measure the height of a tree... of course it may be that the >> thing you're measuring doesn't have >> anything directly below it, in which case you have to measure the slant >> distance. >> >> >>> This worked out pretty good until I found that only short range sensors >>> are >>> cheap :) I'd like to figure out something that would read say a hundred >>> feet ahead. The idea would be to measure the height from an overhang or >>> bridge or opening to ground from a distance. >>> >> >> I think Polaroid do ultrasonic sensors that go up to 30m, but laser would >> be much more reliable. Not sure of >> the costs, though. >> >> >>> I was then thinking of some sort of camera system to capture an image, >>> say 2 >>> per second, but then came the problem of determining automatically what >>> is >>> above and where "ground level" is to that above area. >>> >> >> I'm not sure what your camera would tell you, or why you want more than >> one picture? >> >> >>> So, if you can imagine - in my rather lame ascii art.... >>> >>> Top of overhang >>> -------------- >>> \ >>> | >>> | Heigth to calculate >>> | >>> >>> |---------------------------------------------------------------------- >>> > Point of origin, exact height (give or take) by combination >>> | < -- Distance, 25 - 100 >>> feet --> data entry or perhaps gps (got that idea >>> from >>> looking at gps data >>> | >>> from a remote control helicopter video feed project on the net) >>> / >>> ---------------- >>> Ground >>> >>> I'm sure you can see why the triangle option is very tempting, the >>> problem >>> is distance, doing it without lasers (would suck to have someone walk in >>> front of my new toy and get blinded and sue me because I was >>> experimenting.) >>> >> >> Surveying lasers aren't dangerous to eyesight - they wouldn't be allowed >> otherwise! Imagine roadside >> surveying blinding (even temporarily) a passing driver... I think they >> use Infra Red so you don't even see >> them. >> >> >>> That's why I was thinking images, but I don't even know where I would >>> start >>> research on something like this. >>> >> >> You need to measure a distance - there really is no other way unless you >> know this, otherwise a 10' height at >> 100' looks the same as a 20' height at 200', so anything purely optical >> with no distance measurement won't >> work. >> >> Cheers, >> >> >> Howard Winter >> St.Albans, England >> >> >> > > > -- > http://www.piclist.com PIC/SX FAQ & list archive > View/change your membership options at > http://mailman.mit.edu/mailman/listinfo/piclist -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist