-1

Currently, I am working on a project where I need to calculate the certain Gait Parameters.

Experiment Setup: I have two sensors placed on the shank(near the ankle) of each leg. Each sensor consists of the timestamp, accelerometer, gyroscope and quaternion data with an internal frequency of 100Hz which stores the whole walk data as a comma-separated .txt file.

Now here I want to calculate the stride length, the distance between 2 consecutive heel strike(I can find the heel strike using the gyro data).

Extra parameters: I have subject's height. sensor placement height.

Language used: Python

Ajit Wadalkar
  • 170
  • 8
  • This might be better asked at the mathematics page. And the question is also not completely clear: what data do you really need? If you already know when heels strike the floor, the route followed, the speed, and the time, what prevents you from using this to calculate the distance? – Martijn May 13 '18 at 18:03
  • I will try the mathematics page. I want to calculate the distance between two heel strike. And I can find the time-stamp of heel-strike, time interval between 2 heel strikes. Not the route followed or speed. – Ajit Wadalkar May 13 '18 at 19:05
  • You can start with alreadfy working projects measuring path with dead reconing. – minorlogic May 14 '18 at 07:01

1 Answers1

1

Calculating stride length is difficult without a magnetometer, because you need to ensure that the orientation is known so you can remove the gravity vector before double integrating. There is too much drift in gyroscope data to use it for more than about a minute. Have you done a literature search yet? You will probably want to use a Kalman filter to calculate position. Since the ankle rotates during gait you can not assume the gravity vector will remain in the same position relative to the ankle throughout the gait cycle.

L. Putvin
  • 56
  • 5