Synchronizing Accelerometer Data Received from Multiple Accelerometers and Dynamically Compensating for Accelerometer Orientation

• US 20140236531A1
• Filed: 02/19/2014
• Published: 08/21/2014
• Est. Priority Date: 02/19/2013
• Status: Active Grant

• Current Assignee: Amiigo, Inc.
• Original Assignee: Abaham Carter
• Inventors: Carter, Abaham
• Granted Patent: US 9,936,901 B2
• Time in Patent Office: Days
• US Class Current: 702/141
• CPC Class Codes: A61B 5/1123 Discriminating type of movement, e.g. walking or running A61B5/1116, A61B5/112 take precedence
• CPC Class Codes: G01C 22/006 Pedometers

Abstract

Accelerometer data that is received from multiple accelerometers can be synchronized to identify a known movement. This synchronization occurs in real time as the accelerometer data is received from the multiple devices. The orientation of an accelerometer can be compensated for dynamically. By dynamically compensating for the orientation, the user can wear the accelerometers in various orientations without needing to calibrate a portable computing device to account for a particular orientation. Portable devices can have one or more detachable accelerometers. The accelerometers can be detached and placed at virtually any location of a user'"'"'s body to allow for the detection of custom movements.

Claims

1. A method for synchronizing data received from multiple accelerometers, the method comprising:

receiving multiple streams of accelerometer data generated by multiple accelerometers worn by a user at multiple locations of the user'"'"'s body while the user performs a first activity;

accessing one or more timestamps in each of the multiple streams;

based on the timestamps in each stream, extracting portions of each stream that have a common timestamp indicating that the accelerometer data in the extracted portions was generated during the same time period while the user is performing the first activity; and

using the extracted portions of each stream that share a common timestamp to identify an entry in a database that represents the first activity, the entry including data representing accelerations that are generated at each of the multiple locations of the body when the first activity is performed.

2. The method of claim 1, further comprising:

detecting that one or more of the multiple streams is misaligned with one or more other streams; and

aligning the one or more of the multiple streams with the one or more other streams.

3. The method of claim 2, wherein detecting that one or more of the multiple streams is misaligned with one or more other streams comprises independently processing each of the multiple streams to identify a known movement.

4. The method of claim 3, wherein processing each of the multiple streams to identify a known movement comprises identifying a number of repetitions of the known movement that are represented in each of the multiple streams.

5. The method of claim 2, wherein aligning the one or more of the multiple streams with the one or more other streams comprises time shifting the accelerometer data of at least one of the multiple streams.

6. The method of claim 5, wherein detecting that one or more of the multiple streams is misaligned with one or more other streams comprises:

independently processing each of the multiple streams to identify the occurrence of a first repetition of a known movement in each of the multiple streams; and

wherein the accelerometer data of at least one of the multiple streams is time shifted to cause the occurrence of the first repetition of the known movement in each of the multiple streams to be aligned.

7. The method of claim 6, wherein the accelerometer data of at least one of the multiple streams is time shifted forward and the accelerometer data of at least one of the multiple streams is time shifted backward.

8. The method of claim 1, wherein the multiple streams include streams of accelerometer data generated for multiple axes by the same accelerometer.

9. The method of claim 1, wherein the method is performed by one or more servers which receive the multiple streams of accelerometer data from a portable computing device, the portable computing device having received the multiple streams of accelerometer data from the multiple accelerometers.

10. A method for dynamically compensating for the orientation of an accelerometer, the method comprising:

receiving, by a portable computing device, accelerometer data generated by an accelerometer worn by a user while the user performs a first activity, the accelerometer being worn in an unknown orientation;

accessing a database storing a plurality of feature sets representing accelerometer data that is generated when the first activity is performed, each of the feature sets corresponding to accelerometer data generated when the accelerometer generating the accelerometer data is in a different orientation;

matching the received accelerometer data to a first feature set of the plurality of feature sets corresponding to the first activity; and

based on the matching, identifying that the user is performing the first activity.

11. The method of claim 10, further comprising:

receiving, by the portable computing device, accelerometer data generated by the accelerometer worn by the user while the user performs the first activity, the accelerometer being worn in a second unknown orientation;

accessing the database to match the received accelerometer data to a second feature set of the plurality of feature sets corresponding to the first activity; and

based on the matching, identifying that the user is performing the first activity.

12. A method for identifying a known movement from accelerometer data received from an accelerometer regardless of the orientation of the accelerometer, the method comprising:

receiving accelerometer data generated by an accelerometer worn by a user while the user performs a first activity, the accelerometer being worn in an unknown orientation;

applying a nonlinear dimensionality reduction (NLDR) to the accelerometer data to produce an NLDR feature set that is common to more than one orientation of the accelerometer;

accessing a database storing a plurality of feature sets, each of the stored feature sets comprising a common representation of accelerometer data generated while performing an activity with an accelerometer positioned in various orientations;

matching the NLDR feature set to a first feature set of the plurality of feature sets corresponding to the first activity; and

based on the matching, identifying that the user is performing the first activity.

13. The method of claim 12, wherein the accelerometer data comprises accelerometer data in three axes and the NLDR feature set comprises a two dimensional representation of the accelerometer data in three axes.

14. The method of claim 12, wherein at least one of the plurality of feature sets comprises a common representation of accelerometer data generated while performing an activity with an accelerometer positioned in eight orientations.

15. The method of claim 12, wherein the method is performed by one or more servers which receive the accelerometer data from a portable computing device, the portable computing device having received the accelerometer data from the accelerometer.

16. A portable accelerometer device comprising:

attaching means for attaching the portable accelerometer device to a part of a user'"'"'s body; and

one or more accelerometers that are detachable from the attaching means, the one or more detachable accelerometers being configured to attach to the user'"'"'s body in a plurality of locations while detached from the attaching means.

17. The portable accelerometer device of claim 16, wherein the attaching means comprise one of:

a bracelet;

a shoe clip;

or

an arm band.

18. The portable accelerometer device of claim 16, further comprising:

means for receiving accelerometer data from the one or more accelerometers while the one or more accelerometers are detached; and

means for transmitting the accelerometer data to a portable computing device.

19. A detachable accelerometer comprising:

an accelerometer for transmitting accelerometer data to a portable computing device; and

a connector for attaching the accelerometer to the portable computing device, wherein the connector is removable from the portable computing device to allow the accelerometer to be attached to a user'"'"'s body in a plurality of locations so that movement of the user'"'"'s body at any of the plurality of locations can be identified using the accelerometer data generated while the accelerometer is attached to the location.

20. The detachable accelerometer of claim 19, wherein the accelerometer is configured to perform one or more of:

(1) transmitting accelerometer data directly to a mobile computing device;

or

(2) transmitting accelerometer data to the connector, the connector having circuitry for receiving the accelerometer data and routing the accelerometer data to a mobile computing device.