Calibration of electromagnetic tracking devices

Abstract

Electromagnetic tracking devices are often used to track location and orientation of a user in a virtual reality environment. Their precision, however, is not always high enough due to the dependence of the system on the local electromagnetic field that can be easily altered by many external factors. The purpose of this article is to give an overview of the calibration techniques used to improve the precision of the electromagnetic tracking devices and to present a new method that compensates both the position and orientation errors. It is shown numerically that significant improvements in the precision of the detected position and orientation can be achieved with a small number of calibration measurements to be taken. Unresolved problems and research topics related to the proposed method are discussed.

Electromagnetic Tracking in Medicine—A Review of Technology, Validation, and Applications

“…When new sensors, markers, or fiducials are (custom) built, they must be recalibrated [65]. Finally, calibration also can mean the assessment and compensation of the EM field distortions in a given environment [71]. Regarding the latter, the two fundamental approaches of accuracy improvement include (1) passive protection and (2) active compensation.…”

“…The error compensation can be performed online during the use of the tracking system, based on another independent [31,32] Combination of EM and optical tracking systems Birkfellner et al [17] Using discrete LUT Meskers et al [107] Measurements averaged over time at reference points Perie et al [136] LUT based on plexi phantom Day et al [26] Employing a calibration phantom Interpolation Raab et al [148] First polynomial fit Thormann et al [177] Online correction by Hardy's multiquadric method Himberg et al [57] Multi-sensor data collection with LEGO Ikits et al [62] 4 th order polynomial fit Bryson et al [21] 4 th order polynomial fit Hagedorn et al [50] Delaunay tetrahedralization for rotations Kindratenko et al [71] 3-5 th order polynomial fit Nakamoto et al [124] 0-4 th order polynomial fit Traub et al [181] Interpolation on the top of LUT Fischer [35] Thin-Plate Spline Interpolation, Bernstein-polynoms Kelemen [67] Delaunay tetrahedralization and polynomial fit Extrapolation Kelemen [67] Extrapolation based on global polynomial fit…”

A survey of electromagnetic position tracker calibration techniques

“…in the true space Q and the corresponding tracked locations are stored as {p~} in the tracked space p. Here i=1 ..... n where n is the total number of grid nodes where the measurements were taken. Location error vector v~ at the tracked location p~ can be found as v~=p~-qv A degree r vector polynomial of location p e P that fits the location error v can be formulated as [20,21] …”

Tracking the line of primary gaze in a walking simulator: Modeling and calibration

“…This correction depends on the location portion of magnetic sensor readings. Specifically, to compensate for distortion, the translational components of each raw measurement from the magnetic tracker (x o , y o , and z o , representing the location in three dimensions of the magnetic sensor relative to the transmitter) are transformed with a polynomial distortion correction function (Kindratenko, 1999). The correction consists of three degree-n polynomials in three variables.…”