Is it possible to determine whether a person is sitting or standing by (only) using an Accelerometer? Or is a combination of multiple sensor needed for this?
FYI the Accelerometer is attached to the persons back (I don't know if that makes any difference ...).
Probably not... But if you've got the accelerometer it could be an interesting experiment. It's only going to measure acceleration (changes) and it measures quick movements a lot better than slow movements. It would probably be easy to fool if you were trying and I'm not sure about the accuracy if you're not trying to fool it.
StefaanAvonds:
FYI the Accelerometer is attached to the persons back (I don't know if that makes any difference ...).
Does the sensor necessarily have to be on the person's back? Accelerometers are good at detecting orientation using the fact that the direction of gravity (down) is large and constant unless the subject being tracked is accelerating very quickly. If the sensor were on the person's upper leg there would be a clear difference between the leg being more or less vertical vs more or less horizontal.
If the object to which it is attached is not otherwise accelerating, an accelerometer can report which direction is straight down, relative to its X, Y and Z axes.
An IMU (accelerometer + gyro) output can be used detect movement patterns in order to differentiate between siting down and standing up movements, even if static orientation is the same in both standing and siting positions.
But it's not straightforward. Specially if you expect to gather other data like spine flex..
Pressure sensor on a buttcheek. No need to complicate things.
That is, if you've already made peace with the answer that what you wanted is not possible. If not, I suspect you'll keep asking until you hear what you want to hear.