Performance metrics

• US 8,988,240 B2
• Filed: 12/17/2010
• Issued: 03/24/2015
• Est. Priority Date: 01/15/2009
• Status: Active Grant

• Current Assignee: Avid Sports, Incorporated (Avid Technology Incorporated)
• Original Assignee: Avid Sports, Incorporated (Avid Technology Incorporated)
• Inventors: Burton, Bruce J., Videtich, John, Guthrie, Warren, Johnson, Ryan, Duckworth, Paul
• Primary Examiner(s): Lau, Hoi
• Application Number:
  US12/971,961
• Publication Number:
  US 20110153042A1
• Time in Patent Office: 1,558 Days
• Field of Search: 340/686.1, 340/573.1, 340/572.1, 340/539.13, 340/539.11, 382/103, 382/190, 382/115, 348/157, 348/156, 348/159, 348/564
• US Class Current: 340/686.1
• CPC Class Codes: A63B 2220/12 Absolute positions, e.g. by using GPS
• CPC Class Codes: A63B 2220/40 Acceleration
• CPC Class Codes: A63B 2220/806 Video cameras
• CPC Class Codes: A63B 2225/50 Wireless data transmission, e.g. by radio transmitters or telemetry
• CPC Class Codes: A63B 2244/20 Swimming
• CPC Class Codes: A63B 24/0062 Monitoring athletic performances, e.g. for determining the work of a user on an exercise apparatus, the completed jogging or cycling distance
• CPC Class Codes: A63B 71/0616 Means for conducting or scheduling competition, league, tournaments or rankings
• CPC Class Codes: G01C 23/00 Combined instruments indicating more than one navigational value, e.g. for aircraft; Combined measuring devices for measuring two or more variables of movement, e.g. distance, speed or acceleration
• CPC Class Codes: G06V 40/23 Recognition of whole body movements, e.g. for sport training
• CPC Class Codes: G06V 40/28 Recognition of hand or arm movements, e.g. recognition of deaf sign language static hand signs G06V40/113
• CPC Class Codes: H04N 7/181 for receiving images from a plurality of remote sources

Abstract

A method and system relates to locating or otherwise generating positional information for an object, such as but not limited generating positional coordinates for an object attached to an athlete engaging in an athletic event. The positional coordinates may be processed with other telemetry and biometrical information to provide real-time performance metrics while the athlete engages in the athletic event.

Claims

1. A system for generating performance metrics for a user comprising:

at least two arm sensors operable to sense user arm movement in an x, y, and z direction;

at least one head sensor operable to sense user head movement in an x, y, and z direction;

at least one leg sensor operable to sense user leg movement in at least a y direction;

at least one processor operable to determine performance metrics for the user based on the movement sensed by the arm, leg and head sensors;

wherein at least one performance metric determined by at least one of the at least one processor identifies a number of leg kicks performed by the user between a starting marker and an ending marker based on y values used by the at least one leg sensors to quantify movement in the y direction.

2. The system of claim 1 wherein at least one performance metric determined by at least one of the at least one processor identifies whether left and right arms of the user are performing an alternating swimming stroke based on movement sensed while one of the at least two arm sensors senses user movement of the left arm (referred to as a left arm sensor) and one of the at least two arm sensors senses user movement of the right arm (referred to as a right arm sensor).

3. The system of claim 2 wherein the alternating type swimming stroke is determined in the event a value A is greater than zero, wherein A is determined according to the following formula:

4. The system of claim 3 wherein the alternating swimming stroke is determined by using movement sensed by no more than the left and right arm sensors.

5. The system of claim 1 wherein at least one performance metric determined by at least one of the at least one processors identifies whether the head of the user is orientated to reflect performing one of a head-up type and a head-down type swimming stroke based on movement sensed by the at least one head sensor.

6. The system of claim 5 wherein the head-up type swimming stroke is determined in the event a value H is greater than a threshold and the head-down type swimming stoke is determined in the event the value H is less than the threshold, wherein H is determined according to the following formula:

H=X_H+Z_H;

wherein x_H and z_H are values used by the head sensor to quantify movement in the corresponding x and z directions.

7. The system of claim 5 wherein the heads-up swimming stroke is determined by using movement sensed by no more than the at least one head sensor.

8. The system of claim 1 wherein at least one performance metric determined by at least one of the at least one processor identifies whether left and right arms of the user are performing one of a breast stroke and a butterfly swimming stroke based on movement sensed while one of the at least two arm sensors senses user movement of the left arm (referred to as a left arm sensor) and one of the at least two arm sensors senses user movement of the right arm (referred to as a right arm sensor).

9. The system of claim 8 wherein the breast stroke is determined in the event a value B_a is less than a threshold and the butterfly swimming stoke is determined in the event the value B_a is greater than the threshold, wherein B_a is determined according to the following formula:

B_a=avg(x_L+x_R)−

avg(y_L++y_R)

wherein x_L and y_L are values used by the left arm sensor to quantify movement in the corresponding x and y directions; and

wherein x_R and y_R are values used by the right arm sensor to quantify movement in the corresponding x and y directions.

10. The system of claim 1 wherein at least one performance metric determined by at least one of the at least one processors identifies whether the head of the user is orientated to reflect performing a crawl type swimming stroke based on movement sensed by the at least one head sensor.

11. The system of claim 10 wherein the crawl swimming stroke is determined in the event a value C is greater than a threshold and not determined in the event the value C is less than the threshold, wherein C is determined according to the following formula:

C=y_H/(x_H+z_H);

wherein x_H, y_H, and z_H are values used by the head sensor to quantify movement in the corresponding x, y and z directions.

12. The system of claim 1 wherein at least one performance metric determined by at least one of the at least one processor identifies whether left and right arms of the user are performing one of a breast stroke and a butterfly swimming stroke based on movement sensed by the at least one head sensor.

13. The system of claim 12 wherein:

the breast stroke is determined in the event a best-fit of x_H, z_H, and z_H values used by the head sensor to quantify movement in the corresponding x, y and z directions is with one of a plurality of breast stroke correlation filters; and

the butterfly stroke is determined in the event a best-fit of x_H, z_H, and z_H values used by the head sensor to quantify movement in the corresponding x, y and z directions is with one of a plurality butterfly stroke correlation filters.

14. The system of claim 1 further comprising wherein at least one performance metric determined by at least one of the at least one processor identifies a number of swimming strokes performed by the user between a starting marker and an ending marker based on movement sensed by at least one of the at least two arm sensors and the at least one head sensor.

15. The system of claim 14 wherein the number of swimming strokes is determined by correlating x, y, and z values sensed while the user swims between the starting and ending markers with a reference stroke selected from the x, y and z values, the x, y, and z values being used by the arm sensors to quantify movement in the corresponding x, y, and z directions.

16. The system of claim 15 wherein the correlating of the x, y, and z values with the reference stroke includes filtering the x, y, and z values with the reference stroke to identify amplitude peaks, wherein each amplitude peak correlating with at least a P percentage of the reference stroke is counted as one of the number of swimming strokes.

17. The system of claim 16 wherein the P percentage is at least one of 35% and a value corresponding with a lowest amplitude peak within the greatest ⅔

of the amplitude peaks.

18. The system of claim 17 further comprising at least one of the at least processors characterizing each of the number of swimming strokes as one of matched and irregular, the matched strokes having amplitude peaks of greater than 79% and each irregular stroke having amplitude peaks of between 79% and 35%.

19. The system of claim 16 further comprising characterizing a first stroke occurring proximate the starting marker to be a matched stroke in the event the corresponding amplitude peak is at least 50% and a last stroke occurring proximate the ending marker to be a matched stroke in the event the corresponding amplitude peak is at least 50%.

20. The system of claim 14 wherein the number of swimming strokes is determined by correlating x and y values sensed while the user swims between the starting and ending markers with a reference stroke selected from the x and y values, the x and y values being used by the head sensor to quantify movement in the corresponding x and y directions.

21. The system of claim 20 wherein:

the correlating of the x and y values with the reference stroke includes filtering the x and y values with the reference stroke to identify amplitude peaks, wherein each amplitude peak correlating with at least a 35% percentage of the reference stroke is counted as one of the number of swimming strokes; and

at least one of the at least one processor characterizes each of the number of swimming strokes as one of matched and irregular, the matched strokes having amplitude peaks of greater than 79% and each irregular stroke having amplitude peaks of between 79% and 35%.

22. The system of claim 1 wherein the number of leg kicks is determined by filtering the y values in the frequency domain to identify amplitude peaks, wherein each amplitude peak is counted as one of the number of let kicks.

23. The system of claim 22 wherein the y values are filtered with a bandpass filter to identify the amplitude peaks in the event the y values are characterized as sinusoidal and wherein the y values are filtered with a highpass filter to identify the amplitude peaks in the event the y values are characterized as impulsed.

24. The system of claim 1 wherein at least one performance metric determined by at least one of the at least one processors identifies a length of time taken for the user to swim one length of a pool based on movement sensed by the at least one head sensor.

25. The system of claim 24 wherein the length of time is counted from a start marker to an end marker, wherein the start and end markers are determined based on movement sensed by the at least one head sensor, the start marker corresponding with movement indicating a first push-off event following a first head event, and the end marker corresponding with movement indicating a second push-off event following a second head event.

26. The system of claim 25 wherein the first and second head events are determined in the event at least one of the following conditions occur:

(i) a head dip defined as z_H−

x_H>

T, wherein T is a threshold and x_H and z_H are values used by the head sensor to quantify movement in the corresponding x and z directions;

(ii) a head throw defined as y_H>

T_d, wherein T_d is a dynamic threshold and y_H is a value used by the head sensor to quantify movement in the y direction; and

(iii) a head roll defined as z_H<

T_Z or S number of seconds before the head dip and z_H>

T_Z for the S number of seconds following the same head dip, wherein T_Z is a threshold and z_H is a value used by the head sensor to quantify movement in the z direction.

27. The system of claim 25 wherein the first and second push-off events are determined in the event z_H−

x_H>

T, wherein T is a threshold and x_H and z_H are values used by the head sensor to quantify movement in the corresponding x and z directions.

28. A method of identifying a swimming stroke performed by a swimmer to be one of a crawl, back, breast, and butterfly stroke, the method comprising:

receiving a signal from each of one or more sensors attached to at least one of a head, one or more legs and one or more arms of the swimmer, each signal representing movement of the corresponding sensor relative to an x, y, and z axis;

processing the signals to identify the swimming stroke to be the crawl stroke in the event at least one of;

(i) the signals indicate the head to be orientated head-up and the swimmer'"'"'s arms to be alternating; and

(ii) the signals indicated the head to be orientated head-up and rolling side-to-side;

processing the signals to identify the swimming stroke to be the back stroke in the event the signals indicate the head to be orientated head-down;

processing the signals to identify the swimming stroke to be the breast stroke in the event at least one of;

(i) the signals indicate the swimmer'"'"'s arms to be alternating and the swimmer'"'"'s hand angle moves to a vertical position when swimming; and

(ii) the signals indicate the head to be orientated head-up, the head is not rolling side-to-side, and a stroke pattern best-fits with a known breast stroke pattern;

processing the signals to identify the swimming stroke to be the butterfly stroke in the event at least one of;

(i) the signals indicate the swimmer'"'"'s arms to be alternating and the swimmer'"'"'s hand angle remains in a flat position when swimming;

(ii) the signals indicate the head to be orientated head-up, the head is not rolling side-to-side, and a stroke pattern best-fits with a known butterfly stroke pattern; and

processing the signals to identify a number of leg kicks performed by the swimmer between a starting marker and an ending marker based on y-axis values generated with the at least one leg sensor to quantify movement in a y direction.

29. A method of counting a number of strokes performed by a swimmer from a movement signal generated by at least one of an arm sensor, a leg sensor and a head sensor during a swum lap, the method comprising:

identifying a dominate stroke period of the swum lap from the movement signal;

selecting a portion of the movement signal coinciding with one of the dominate periods to be a first reference stroke;

filtering the movement signal based on the first reference stroke to identify a number of potential strokes;

determining the number of strokes based on the identified number of potential strokes; and

determining a number of leg kicks performed by the swimmer between a starting marker and an ending marker based on y values generated with the leg sensor to quantify movement in a y direction.

30. The method of claim 29 further comprising:

averaging portions of the movement signal corresponding with two or more of the potential strokes to generate a second reference stroke;

filtering the movement signal based on the second reference stroke to identify a number of final strokes; and

determining the number of strokes to be the number of final strokes having amplitudes greater than a threshold.

31. The method of claim 29 further comprising centering phase of the first reference stroke to a phase of a first stroke of the swum length prior to filtering the movement signal to identify the number of potential strokes.

32. The method of claim 31 further comprising the centering including aligning the reference stroke relative a beginning of the first stroke, the beginning of the first stroke being determined with an activity-detection filter that filters a region of the first stoke to identify a transition point from no activity to activity, the transition point being the beginning of the first stroke.

33. A method of counting a number of kicks performed by a swimmer based on a movement signal generated by at least one leg sensor during a swum lap, the method comprising:

filtering the movement signal with a first plurality of bandpass filters to generate a corresponding first plurality of filtered movement signals;

measuring an amplitude variance associated with each of the first plurality of filters;

in the event the amplitude variance is less than a threshold, filtering the movement signal with a second filter to produce a second movement signal and counting the number of kicks to correspond with a number of peaks identified within the second movement signal, the second filter being different from the plurality of filters; and

in the event the amplitude variance is greater than the threshold, identifying one of the first plurality of bandpass filters matching a dominate frequency of the movement signal, the bandpass filter matching the dominate frequency of the movement signal being the one of the first plurality of bandpass filters generating the greatest amplitude output, and counting the number of kicks to correspond with a number of peaks identified within the corresponding one of the first plurality of filtered movement signals.

34. The method of claim 33 further comprising selecting the first plurality of filters to be lowpass and the second filter to be highpass.

35. A method of determining a timestamp to reflect a point in time when a swimmer executes a push-off based on a signal that three-dimensionally represents movement of the swimmer while swimming towards a wall prior to executing the push-off and way from the wall after executing the push-off, the method comprising:

processing the signal to identify a first stroke occurring after the swimmer executes the push-off event;

identifying a portion of the signal associated with swimmer movement prior to the first stroke for occurrence of a push-off event;

in the event a single push-off event is identified, determining the timestamp to coincide with the single push-off event;

in the event multiple push-off events are identified,

(i) determining the timestamp to correspond with an earliest one of the multiple push-off events occurring within a push-off search region measured from occurring of a single head event,

(ii) determining the timestamp to correspond with an earliest one of the multiple push-off events occurring within an earliest one of multiple push-off search regions measured from occurrence of a corresponding multiple of head events, and

(iii) determining the timestamp to correspond with an earliest one of the multiple push-off events occurring outside of any push-off search regions measured from occurrence of a corresponding head event; and

determining a number of leg kicks performed by the swimmer between a starting marker and an ending marker based on y values generated with a leg sensor worn on the swimmer to quantify movement in a y direction.

36. The method of claim 35 further comprising determining push-off events by high-pass-filtering X and Z axis data of the signal, subtracting Z axis data from X axis data, and finding peaks above threshold, each peak above the threshold being one push-off event.

37. The method of claim 35 further comprising determining head events to be any one of a head throw, a head dip, and a head roll wherein:

the head dip is found by low-pass-filtering X and Z axis portion of the signal, subtracting Z from X, and finding a peak above a threshold, each peak being one head dip;

the head roll is found by identifying a minimum in a low-pass-filtered Z axis portion of the signal; and

the head throw event is generated by squaring a low-pass-filtered Y axis portion of the signal to find peaks above a threshold, each peak being one head throw.

38. A non-transitory computer-readable medium having a plurality of non-transitory instructions operable with a processor to facilitate counting a number of kicks performed by a athlete based on a movement signal generated by at least one leg sensor worn on the athlete, the method comprising:

filtering the movement signal with a first plurality of bandpass filters to generate a corresponding first plurality of filtered movement signals;

measuring an amplitude variance associated with each of the first plurality of filters;

in the event the amplitude variance is less than a threshold, filtering the movement signal with a second filter to produce a second movement signal and counting the number of kicks to correspond with a number of peaks identified within the second movement signal, the second filter being different from the plurality of filters; and

in the event the amplitude variance is greater than the threshold, identifying one of the first plurality of bandpass filters matching a dominate frequency of the movement signal, the bandpass filter matching the dominate frequency of the movement signal being the one of the first plurality of bandpass filters generating the greatest amplitude output, and counting the number of kicks to correspond with a number of peaks identified within the corresponding one of the first plurality of filtered movement signals.