We’re all still getting calibrated on sensors, and we can sometimes get tripped up on terminology. In discussions about their sensor fusion IP, Movea explicitly included some definitions to make sure we were aligned.
Most prominent is their finessing of the mythical 9-axis or 9-DoF sensor. They do this by distinguishing between true degrees of freedom and axes. Spatial motion involves only six true degrees of freedom: 3 dimensions linear and three dimensional rotational. That’s all there is. The problem is that when accelerometers (which handle 3D linear) and gyroscopes (which handle 3D rotational) are joined by magnetometers, which also position in 3 dimensions, many marketers blithely add the degrees of freedom together to get 9 of them.
The problem is, the magnetometer’s three degrees of freedom overlap those of the other two sensors. We haven’t magically entered a new world that has three completely new ways of moving; we’re just getting a new view of the existing six degrees of freedom.
Movea is more permissive with the term “axis,” allowing each sensor its own axes and making them additive. So an accelerometer/gyroscope/magnetometer combo, in their parlance, has six degrees of freedom and nine axes.
Other lexical distinctions they provide:
– Axial sensors, which provide, essentially, a vector result, with a value for each axis; versus scalar sensors, like pressure sensors, which simply provide one number as a result
– Motion sensing categories, from detection, which is Boolean – either there is motion or there isn’t – to classification, which categorize movements like gestures into discrete buckets, to estimation, where a value from a continuous range is reported – part of what separates full motion processing from gesture recognition.