Determination of

rotation angles and geographic orientations in geomagnetic observatory

instruments using digital camera images

Merényi, L. (1), Hegymegi, L. (1), Domján, Á.


(1) Mingeo Ltd.,

Budapest, Hungary


At some types of

geomagnetic observatory measurements, it is crucial to accurately

measure the relative rotation angles and/or geographic orientations of

the magnetometer sensors. A typical example is the absolute

measurements using theodolite and a faraway geographic reference point

called MIRA. The required accuracy and resolution of such angle and

orientation measurements is very high, while the measuring device

should be non-magnetic, that complicates the instrument design and the

realization of the measurement.

Using appropriate digital image processing methods on images taken by

digital cameras, the movements of the recorded objects, including the

rotation of the objects can be estimated. We tested some possible

configurations and some digital image processing methods to measure

the relative rotation of a magnetometer sensor holder, and to relate

these relative measurements to reference directions in order to get

geographic orientations.


In this paper we present our tested methods and the results, including

the estimated resolution and accuracy. We also present some ways how

the method could be employed in absolute measurements.


There are some considerable advantages of the suggested image-based

procedure. Putting the camera away from the magnetic sensor, its

magnetic effect can be minimized. Moreover, there are more

possibilities to relate the rotation over the vertical axis to the

true North (e.g. in declination measurements), and to relate the

rotation over the horizontal axis to the vertical direction (e.g. in

inclination measurements). We tested the use of pictures taken from

sunlight shadow patterns in a horizontal plane, and the use of

pictures taken from free hanging strings. These pictures can give

natural horizontal and vertical reference directions for absolute

measurements. We tested some other methods to relate the measurements

to true North in the absence of the sunlight.


Limitations and errors due to the limited resolution and pixel noise

of the digital images can be efficiently reduced with some simple

digital computing methods. There are also some options to check and to

reduce the possible errors due to optical distortion of the camera, to

the non-ideal camera positions and to the long term small movements of

the parts of the camera measuring system.