Very noisy Roll Pitch Yaw data
YunHB opened this issue · comments
Hello. Thank you for those nice libraries.
I'm trying to use those libraries for MARG sensor fusion, and when I logged those data I got data with huge noise.
I am thinking of calibration of sensors and introduction of low-pass filters for sensor data as a solution to the noise.
May I get some hints with your opinion?
I used the c headers and sources.
those pictures are compared shots with commercial IMU module.
It looks like your Euler angles are garbage, not noisy. I suggest you format your sensor data into a CSV file to be processed by simple_example.py. This will plot gyroscope, accelerometer, and Euler angles side-by-side as shown below. This plot may offer some insight as to what is going on.
It looks like garbage, but with same imu, same sensors, other algorithm got very accurate euler angle data.
And the difference I found between those two algorithm is existence of LPF on sensor data.
I will try your suggestion, but I think it won't make this better.
My suggestion will not make it better because I did not suggest a solution. I suggested that you provide plots that would hopefully allow us to see what the problem is.
OH that was my misunderstanding, sorry. I applied LPF and HPF on sensor data, and I got a better plot. Maybe the sensors have too much noise or I treated them in unstable condition.
If you are not going to engage with this support then please close the issue.