Triangulation in three dimensions

This is an old revision of this page, as edited by EdwardLane (talk | contribs) at 14:26, 24 May 2011 (grasped some of the meaning - tidied up a bit based on that). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

In vector analysis, triangulation is a method of finding points in three dimensional spaces using distances, angles and vector functions such as magnitude, dot product and cross product. Among its uses are in surveying, navigation, and astronomy.

This article describes a method for determining the coordinates of the point where three lines meet, given the scalar lengths of the lines and the coordinates of their bases. First treat these three lines as if they are the radii of three spheres of known centers (these known centres being the coordinates of the known end of each line), this method can then be used to calculate the intersection of the three spheres if they intersect. In the event that the three spheres don't intersect, this method obtains the closest solution to the axis of symmetry between three spheres.

Development

File:Triangulation illust 02.gif
Figure 1. The apex and its mirror reflection about the plane of ABC precipitate D and D'.

Three sticks of known lengths AD, BD, CD are anchored in the ground at known coordinates A, B, C. This development calculates the coordinates of the apex where the other ends of the three sticks will meet. These coordinates are given by the vector D. In the mirror case, D' is sub-apex where the three sticks would meet below the plane of A, B, C as well.


By the law of cosines,

 
 
 







File:FacesABD ACD BCD 2.gif
Figure 2. The normals are dropped on the sides from the apex and their intersections with AB, AC and BC are determined.


The projection[1] of AD onto AB and AC, and the projection of BD onto BC results in,

 
 
 



 
Figure 3. The red normals intersect at a common point.


The three unit normals to AB, AC and BC in the plane of ABC are:


 


 


 


Then the three vectors intersect at a common point:

 



Solving for mAB, mAC and mBC

 


Spreadsheet formula

A spreadsheet command for calculating this is,

PRODUCT(PRODUCT(MINVERSE(PRODUCT(TRANSPOSE H, H)), TRANSPOSE H), g)

An example of a spreadsheet that does complete calculations of this entire problem is given at the External links section at the end of this article.

The the matrix H and the matrix g in this least squares solution[2] are,

 


Alternatively, solve the system of equations for mAB, mAC and mBC:

 

The unit normal to the plane of ABC is,

 


Solution


 



 



where

 
 
 



Condition for intersection

If AD, BD, CD are assigned according to the arrangement,

 

Then AD, BD, CD intersect if and only if,

 
 
 


Viz, if
AD=rA=radius of sphere centered at A,
BD=rB=radius of sphere centered at B, and
CD=rC=radius of sphere centered at C,
such that,

 

then the three spheres intersect if and only if,

 
 
 


Decoding vector formulas

 
 
 


 
 
 


 
 
 


 
 
 


 


 


 


 
 
 


 
 
 


 
 
 


 
 
 


 
 
 


 
 
 


 
 
 


 
 
 


The equation of the line of the axis of symmetery of 3 spheres is,

 


 


 


 


Example

File:3 spheres2abcplane.gif
Figure 4. The data is input into the equations to obtain the solution. The shaded area is the plane of ABC. A, B and C are the centers of each of the three spheres.



File:3 spheres2ABCcenters.gif
Figure 5. Showing points D and D' as the result of 3 intersecting spheres at centers A, B, C from data given in Figure 4. The line adjoining D and D' imbedded in the interiors of all three spheres is the axis of symmetry of the three spheres.




See also

References

  1. ^ Borisenko, A. I. and Tarapov, I. E., (1968) "Vector and Tensor Analysis", General Publishing Company, p. 6. ISBN 0-486-63833-2
  2. ^ Leon, Steven J. (1980) "Linear Algebra", Macmillan Publishing Co., Inc., p. 152. ISBN 0-02-369870-5