I want to use MEMS accelerometer to align panoramas obtained from 360-panoramic camera. I need g-vector orientation measurement precision about 0.1 degrees. I know that bias instabilty affects precision, but I don't know how to convert values specified in datasheet to actual measurement error in degrees.
Are there any random fluctuations in accelerometer bias at a constant temperature? If it exist how much is it in degrees? I don't want to periodically recalibrate accelerometer.
Is there any application note for BMI160?
Do you mean you align panoramas obtained from 360-panoramic camera only with accelerometer , without gyroscope? If yes, could we know how you align it?
Yes, only with accelerometer. This is horizontal alignment.
I find the orientation of a g-vector in IMU coordinate system by three components x, y, z. Of course the camera is motionless at that moment. By orientation I mean rotation matrix. Then I rotate the panorama stretched over the sphere and for that I use rotation matrix obtained before.
The unit of acceleration is g, i.e. 9.8m/s², which cannot be converted into degrees unit. The unit of gyroscope is degrees per second.