Techniques for determining orientation of a three-axis accelerometer

1. An apparatus comprising:

a memory;

a communications module to receive accelerometer data;

a processor coupled to the memory and the communications module, the processor operative to:

determine an acceleration due to gravity from the received accelerometer data on each axis x, y, z of a three-axis on-body accelerometer regarding forward motion by a wearer of the accelerometer;

associate the x axis with the direction of gravity by rotating the x axis to align with the direction of gravity;

identify the acceleration due to forward motion by rotating the y and z axes about the x axis until a projection of forward motion between y and z is minimized;

associate the y axis with the direction of forward motion by rotating the y axis to align with the direction of forward motion;

identify the sideways direction; and

associate the z axis with the sideways direction.

2. The apparatus of claim 1 , wherein the processor is operative to:

select the rotation angles that cause the variance of the y axis to be greater than the variance of the z axis; and

examine time domain characteristics of the cross-correlation of the x and y axes; and

associate the y axis with the direction of forward motion and z axis with the sideways direction by rotating the y axis to align with the direction of forward motion.

3. The apparatus of claim 1 , wherein the processor is further operative to perform activity recognition with the associated axes.

4. The apparatus of claim 1 , further comprising the three-axis on-body accelerometer.

5. The apparatus of claim 1 , wherein the processor is further operative to associate the x axis with the direction of gravity by calculating the angle (a x ) that the x axis makes with the direction of gravity and storing a x in the memory.

6. The apparatus of claim 5 , wherein the processor is further operative to identify the acceleration due to forward motion by:

removing components of the acceleration due to gravity from each axis;

removing the vector projection of the motion in the direction of gravity; and

identifying the axis having the greatest energy as the forward direction.

7. The apparatus of claim 6 , wherein the processor is operative to identify the axis having the greatest energy as the forward direction by:

combining axes having a correlation higher than a threshold;

identifying the axis or combined axes having the highest energy as the forward direction; and

examining time domain characteristics of the cross-correlation of the x and y axes; and

associating the y axis with the direction of forward motion by calculating the angle (a y ) that the y axis makes with the direction of forward motion and storing a y in the memory.

8. The apparatus of claim 7 , wherein the processor is operative to:

identify the sideways direction by calculating the cross-product of the forward direction and the downward direction; and

associate the z axis with the sideways direction by calculating the angle (a z ) the z axis makes with the sideways direction and storing a z in the memory.

9. A method comprising:

receiving accelerometer data regarding forward motion by a wearer of a three-axis on-body accelerometer;

measuring an acceleration due to gravity on each axis x, y, z of the accelerometer;

associating the direction of gravity with the x axis;

rotating the x axis to align with the direction of gravity;

identifying the acceleration due to forward motion;

rotating the y and z axes about the x axis until a projection of forward motion between y and z is minimized;

associating the direction of forward motion with the y axis;

identifying the sideways direction;

associating the sideways direction with the z axis

rotating the y axis to align with the direction of forward motion and the z axis with the sideways direction; and

performing an activity using the associated axes.

10. The method of claim 9 , further comprising:

selecting the rotation angles that cause the variance of the y axis to be greater than the variance of the z axis; and

examining time domain characteristics of the cross-correlation of the x and y axes.

11. The method of claim 9 , further comprising calculating the angle (a x ) that the x axis makes with the direction of gravity and storing a x .

12. The method of claim 11 , further comprising:

removing components of the acceleration due to gravity from each axis;

removing the vector projection of the motion in the direction of gravity;

identifying the axis having the greatest energy as the forward direction; and

calculating the angle (a y ) that the y axis makes with the direction of forward motion and storing a y .

13. The method of claim 12 , further comprising:

combining axes having a correlation higher than a threshold;

identifying the axis or combined axes having the highest energy as the forward direction; and

examining time domain characteristics of the cross-correlation of the x and y axes; and

calculating the angle (a y ) that the y axis makes with the direction of forward motion and storing a y .

14. The method of claim 13 , further comprising:

calculating the cross-product of the forward direction and the downward direction; and

calculating the angle (a z ) the z axis makes with the sideways direction and storing a z .

15. An article comprising a non-transitory computer-readable storage medium containing instructions that when executed by a processor enable a system to:

receive accelerometer data regarding forward motion by a wearer of a three-axis on-body accelerometer;

measure an acceleration due to gravity on each axis x, y, z of the accelerometer;

associate the direction of gravity with the x axis;

rotate the x axis to align with the direction of gravity;

identify the acceleration due to forward motion;

rotate the y and z axes about the x axis until a projection of forward motion between y and z is minimized;

associate the direction of forward motion with the y axis;

identify the sideways direction;

associate the sideways direction with the z axis

rotate the y axis to align with the direction of forward motion and the z axis with the sideways direction; and

perform an activity using the associated axes.

16. The article of claim 15 , further comprising instructions that when executed enable the system to:

select the rotation angles that cause the variance of the y axis to be greater than the variance of the z axis;

examine the time domain characteristics of the cross-correlation of the x and y axes.

17. The article of claim 15 , further comprising instructions that when executed enable the system to:

calculate the angle (a x ) that the x axis makes with the direction of gravity and storing a x .

18. The article of claim 17 , further comprising instructions that when executed enable the system to:

remove components of the acceleration due to gravity from each axis;

remove the vector projection of the motion in the direction of gravity;

identify the axis having the greatest energy as the forward direction.

19. The article of claim 18 , further comprising instructions that when executed enable the system to:

combine axes having a correlation higher than a threshold;

identify the axis or combined axes having the highest energy as the forward direction; and

examine the time domain characteristics of the cross-correlation of the x and y axes; and

calculate the angle (a y ) that the y axis makes with the direction of forward motion and storing a y .

20. The article of claim 18 , further comprising instructions that when executed enable the system to:

calculate the cross-product of the forward direction and the downward direction; and

calculate the angle (a z ) the z axis makes with the sideways direction and storing a z .