bmswgnp / imumargalgorithm30042010sohm

Automatically exported from code.google.com/p/imumargalgorithm30042010sohm
0 stars 0 forks source link

Not an issue #3

Open GoogleCodeExporter opened 8 years ago

GoogleCodeExporter commented 8 years ago
Dear Sirs,

I have read your great report, and I have noticed that you have mentioned and 
also there are lot of references regarding motion of a human body. I would like 
to build a 9dof imu to analyse human running gait, I have results with 3 axis 
accelerometer, but I am not satisfied. I would like to ask you for your opinion 
if this technique that you are describing is the right way to analyse human 
moving. 
I was thinking about to buy ready made boards sold by Sparkfun,..., which 
integrates accelerometer, gyro and compass, all connected to i2c, but I'm 
worried about i2c timing- sensor data would not be captured at the same moment 
sychronously because of nature of serial communication. I have read that 
performance rapidly gains with sampling frequency and then remains almost equal 
for sampling rate of 500hz or 50hz, so at lower frequency also different 
timestamps of measured data would give less error (my way of thinking). What 
would be the adequate frequency for recording a running human ?

Anyway, thank you for great explanation in the paper.

Original issue reported on code.google.com by marko.bu...@gmail.com on 21 Feb 2011 at 7:45