animalTrack (version 1.0.0)

calibrate.axis: Scale and center an accelerometer/magnetometer axis.

Description

Scale and center x using the minimum and maximum values of the sensor. scale value indicates the maximum value x will have after being scaled (i.e. values will range -scale to +scale).

Usage

calibrate.axis(x, max, min, scale = 1)

Arguments

x

a numeric vector of axis values.

max

maximum uncalibrated value of the axis. This should be measured empirically through a series of calibration rotations. For a magnetometer axis, max should be the value of the Earth's "total field" at the locality (i.e. when the axis is aligned with the Earth's total field vector). The min value would conversely be the value when the axis is pointed in the opposite direction. The same logic applied if using accelerometer axis, but instead of the Earth's magnetic field, the gravity vector of the Earth is used (i.e. vector pointing toward the center of the Earth).

min

see max. value when axis is pointed opposite of reference vector (e.g. Earth magnetic or gravity field).

scale

the maximum value x will have after being scaled (i.e. values will range from -scale to +scale. Defaults to -1 to +1.

Value

x.

Warning

If any dynamic acceleration is present during the calibration procedure or track, it will result in an erroneous calibration. This can be avoided by keeping the sensor still and by pausing during rotations. Once the calibration constants are known for the instrument, they can be applied to the raw signal from the deployment on the animal. But remember, the raw accelerometer signal will contain both static and dynamic acceleration, so the accelerometer signal must be filtered using a low-pass filter to extract the static signal.

Details

This function is useful for calibrating the axes of magnetometers/accelerometers and correcting for hard-iron distortion. Centering and scaling each axis will standardize and essentially calibrate each axis so that the simulaneous measurements of each will be trigonometrically accurate. This function effectively deals with hard iron distortion problems as well as small misalignments between axes. However, there are other factors that can cause discontinuity (e.g. temperature, electronic noise, etc.) that are not accounted for in this function. For a more complete discussion of dealing with compass errors see Caruso (2000).

References

Caruso, M. J. (2000), Applications of magnetic sensors for low cost compass systems. Proc. Position Location and Navigation Symp., IEEE 2000, Mar. 2000, pp. 177-184

See Also

calibrate.axes

Examples

## Import the missionbay_calib2 dataset. There is a calibration sequence 
## at Mission Bay, CA. Use help(missionbay_calib2) for info.

data(missionbay_calib2)
mx <- missionbay_calib2$mx_raw
my <- missionbay_calib2$my_raw
mz <- missionbay_calib2$mz_raw

ax <- missionbay_calib2$ax_raw
ay <- missionbay_calib2$ay_raw
az <- missionbay_calib2$az_raw
## Scale and center the magnetometer axes. Also, make sure that each axis 
## is in the NED (north-east-down) orientation (this is the reason for the
## negative signs).

Mxscaled <- calibrate.axis(mx, max(mx), min(mx))
Myscaled <- -calibrate.axis(my, max(my), min(my))
Mzscaled <- -calibrate.axis(mz, max(mz), min(mz))

## Scale and center the accelerometer axes. Also, make sure that each axis 
## is in the NED (north-east-down) orientation.

Axscaled <- -calibrate.axis(ax, max(ax), min(ax))
Ayscaled <- calibrate.axis(ay, max(ay), min(ay))
Azscaled <- calibrate.axis(az, max(az), min(az))

## Compare
par(mfrow=c(2,3))
plot(mx,type="l",col="blue",xlab="time (s)",ylab="bits",main="Mx Raw",lwd=3)
plot(my,type="l",col="blue",xlab="time (s)",ylab="bits",main="My Raw",lwd=3)
plot(mz,type="l",col="blue",xlab="time (s)",ylab="bits",main="Mz Raw",lwd=3)
plot(Mxscaled,type="l",col="red",xlab="time (s)",ylab="bits",main="Mx Calibrated",lwd=3) 
plot(Myscaled,type="l",col="red",xlab="time (s)",ylab="bits",main="My Calibrated",lwd=3) 
plot(Mzscaled,type="l",col="red",xlab="time (s)",ylab="bits",main="Mz Calibrated",lwd=3)