hi @ all !
hoping that someone could help me out with this IMU issue.
short explanation: i started to study art again, and im at the moment trying to make a piece with oculus rift and my FreeIMU.
in short doing something with a oculus and a (any) elevator.. as soon as the elevator starts to move you see a virtual room with the occulus, when the elevator reaches its destination the occulus turns black again.
i therefor need to know when does the elevator start to move and when does it stop, i dont need to know how far it moved or anything precise.. just the (if possible, instant) moment of the starting and ending motion eg. acceleration/decceleration. now, i dont think that i can do this whole thing only by using the accelerometer inside the backpack the user is wearing, i need to implement the barometerdata,.. not that every small bobbing screws up everything.
so my question would be.. does anyone know how i could go about this ? how can i fuse the barometer- with the accelerometerdata ? the barometer has range of about 40cm +/- therefor its not good enough to recognize the impuls of the movement/the start of the elevator. or is there anotherway ?
would really appreciate your opinion about this!
thanks a lot!
numu
IMU Question - Barometer Accelerometer DataFusion
-
- One Eyed Hopeful
- Posts: 15
- Joined: Mon May 21, 2012 5:00 pm
-
- Petrif-Eyed
- Posts: 2708
- Joined: Sat Sep 01, 2012 10:47 pm
Re: IMU Question - Barometer Accelerometer DataFusion
A digital barometer can measure the difference between standing and sitting. It can be used much like a GPS to compensate for drift. However, to determine when an elevator starts or stops, you should be able to just look for positive and negative variations in measured gravity, much like how fall detection works in laptops, and in fall detectors that are worn by fall-prone elderly people.
You just need to threshold changes in measured gravity, looking for vertical increase or decrease that exceeds a pre-defined threshold (determined by elevator start and stop measurements).
To prevent drift caused by noise, you need to force speed to zero when below a minimum movement threshold. That prevents measured position from drifting while the Rift is stationary, but has a side-effect of not being able to measure very slow movements that overlap sensor noise and digital quantization errors. Temperature can also be a factor that limits the minimum threshold.
FYI, I have a digital barometer that I plan to fuse with my GPS and accelerometers, when I get caught up on my currently massive workload (only getting online a few times per week to check my email)...
You just need to threshold changes in measured gravity, looking for vertical increase or decrease that exceeds a pre-defined threshold (determined by elevator start and stop measurements).
To prevent drift caused by noise, you need to force speed to zero when below a minimum movement threshold. That prevents measured position from drifting while the Rift is stationary, but has a side-effect of not being able to measure very slow movements that overlap sensor noise and digital quantization errors. Temperature can also be a factor that limits the minimum threshold.
FYI, I have a digital barometer that I plan to fuse with my GPS and accelerometers, when I get caught up on my currently massive workload (only getting online a few times per week to check my email)...
This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.
-
- One Eyed Hopeful
- Posts: 15
- Joined: Mon May 21, 2012 5:00 pm
Re: IMU Question - Barometer Accelerometer DataFusion
thanks geekmaster!
id be very interested in your solution with the gps/barometer/acc. fusion, i just dont get it, guess id need to learn the kalman filter to finally do it myself, problem is that im a real noob in coding and math as well : )
about the accelerometer problem you pointed out, it would accually be no problem because im using the rift only for the headtracking, in the backpack there would be the FreeIMU wich has a great calibrated accel. built in.
i understood your idea with the gravity, ill definetly check this out and ask some programmerfriends how to go about this!
cheers!
numu
id be very interested in your solution with the gps/barometer/acc. fusion, i just dont get it, guess id need to learn the kalman filter to finally do it myself, problem is that im a real noob in coding and math as well : )
about the accelerometer problem you pointed out, it would accually be no problem because im using the rift only for the headtracking, in the backpack there would be the FreeIMU wich has a great calibrated accel. built in.
i understood your idea with the gravity, ill definetly check this out and ask some programmerfriends how to go about this!
cheers!
numu