map the profile of an irregularly shaped object

1mm!?!

Friend,

(Assumed: outdoors, choice of time, have fixed-focus (known focal length) film camera, can't get closer than 100 m...2-d profile only.)

How about?

Take a picture of it.

Profile the picture.

Scale up.

Repeat to validate.

Plane assembly is confronted with similar problems as far as I know "Zeiss" offers an optical system for measuring big structures.

The basic method would be to use a precise angular measuring device in 2 planes (horizontal and vertical) and a measuring base quite big and make 2 series of measurements and then solve the equations in order to obtain the distances and the space coordinates of object points. I am not sure the uncertainty you expect could be obtained.

Use a 1 second theodolite with an accurate EDM. A distortion free camera lens can also be used. The photo can then be plotted.

A 1" angle gives an error at 100 m of about 3mm !

I am always surprized when values are mentioned without a simple check for their validity.

Coupled with an accurate EDM this error can be reduced to .5mm. Double axis shot fom theo reduces error to 1.5mm.

The Mekometer (an EDM) is accurate up to 1mm at a distance of 40 kilometres.

I am always surprised when validity is not checked with a simple and informed knowledge base.

I am not accustomed to acronyms what is a EDM?

How a double axis shot reduces the error to its half?

Fortunately I do not know all so that I am always ready to learn more. I hope will be so kind to give some quantitative explanations.

EDM.......Electronic Distance Measure (Infra-Red laser coupled with galium arsenide chip).

Double axis shot means turning the the optical assembly of a theodolite 180' and calculating (ostensibly) differance in measure. This reduces the inherant mechanical error by 1/2.

One may also use ECDS........comparative measure using two or more theodolites.

Or........photogrammetrically plotting from a photo taking into account lens distortion (which can be factored in.....similar to earth curvature algorithm).

Could you please be so kind and give a more detailed description of the procedure since I have the feeling that there is some where(according to what I know) a contradiction. I would like to understand correctly what you mean.

Thanking you in advance

Which procedure? Using a theodolite or an edm or a plotter? Or the principals of mechanical accuracy?

The short answer is that it's simple trig. coupled with high precision opto/digi/mechanics.

Yeah, that's right and holding 1 second of arc is going to be very difficult.

I would suggest using a laser scanner but breaking down the area scans into something managable. But even then you aren't going to get the accuracy you require.

What do you mean 'holding one second of arc'....a theodolite is fixed on a tripod or likewise base. Barring there are no earthquakes or tremors the readings will be consistant.

Don't you have adequate instruments for mapping land?

If no here is a easy solution.

make a metal or wood pole and fix a meter size circle on it. post it on every corner of land for each reading, similarly make 2 small metal rod and make them sliding on a meter scale.and fix a small Known size circle on top of rod of both rod.take a table and put this on table.Like a below picture, I hope you understand what to do next. If not pls reply.

The text of the drawing is too small to magnify and read, so let's see if I understand this:

On each of the boundary intersections of the land, except one, I'm going to stick a vertical pole into the ground. On each pole I will hang a circle fabricated of wire that is 1 metre in diameter; all of the circles will all be at the same height above the ground; and each circle will be situated such that, looking through it, I will be looking directly through the circle's center and directly at the one land boundary intersection at which I did not place a pole and a circle.

Next, at and from that intersection without a circle, I am going to measure the distance, one by one, to each of the circles I planted; I will do it like this:

Using a small, wire-fabricated circle of known diameter, and placing my eye at the Zero graduation mark on a metre stick, I will hold the small circle close to my eye and sight through it towards the center of one of the 1-metre-diameter circles. As I sight through the circle, I will then slide the little circle away from me along the metre stick until the circumference of the little circle and the circumference of the distant, 1-metre circle appears to coincide. I will then read the metre stick graduation where the little circle is (where it is when the circles appear to coincide), and write that down. I will turn, in sequence, and repeat the same process for all the 1-metre circles until I have gotten a reading (on the metre stick) for each and every one of them.

Next, on a large piece of paper laid on a table top (or on the table top itself), near a corner of the paper, I will first draw a point that represents the point (the boundary intersection) from which I was sighting. Then, from that point on the paper, and using the metre stick as a straight edge, I will place the metre stick down flat on the paper so that it points directly from the point I drew towards the center of the first, one-metre circle I took a sighting on, and draw a staight line across the paper...as if drawing it towards the center of the distant, 1-metre circle. After that, one by one, I will do the same thing for the other, distant one metre circles. (When drawing these straight lines, I will make sure that the length of each is longer than the highest sighting measurement I wrote down earlier.)

At this time, on the paper I will see lines radiating out from a single point towards the far edges of the paper; if the land is a triangle there will be two lines; if a rectangle...three lines; and so on.

Next, using the sighting distances I wrote down before, for each of the lines on the paper, I will measure along the line from the point where the lines intersect, and will make a tick mark at the distance originally written down for that line. After I do this for each of the lines and corresponding line measurements I wrote down, I will then use the metre stick to draw straight lines connecting the tick marks.

This will give me a scaled plot of the the land and, if I want, I can write the measurement of each line on the plot using the measurements I wrote down. At this point, I will measure the lines connecting the tick marks and write those distances on the plot as well. I will now have a scaled "map" with the scaled length of each plotted land boundary marked on it.

Finally, letting the circle (the little one's and the big ones') diameters represent the height of two congruent triangles, I will calculate the scale of the "map" and write that down on the map. Using the scale figure, I can then determine the actual, land distance of each of the land boundaries and write these next to their corresponding lines on the map.

Is this what you meant?

Thanks for the handy tip...even if it won't help much to answer the original post, where the profile plotting tolerance is only 1mm. Will it?

You have re-invented an ancient (and, incidentally,still used....eg astrolabe, sextant etc) technique that once built the pyramids. The instrument that Rakesh describes is called an alidade...a forerunner to modern optical measuring instruments. Some types are very accurate.

If the object were perfectly 2D, achieving 1mm accuracy from 100 meters would be difficult. With a 3D object, even if all you want is a 2d rendition (a cross section) 1mm accuracy would be very difficult. You might look into the 3D scanners used to scan things like boat hulls. With some of these, you set up several laser units on tripods and points are captured through triangulation in space. This small system will scan large objects, but not from 100 meters away. Janicki industries has been doing this sort of thing for a long time, and may be able to steer you in the right direction. They can go from drawing to large object (via huge CNC milling machines) or from object, to drawing, (to refinements, probably) to replica.