Systems and Methods of Swimming Analysis

Abstract

Systems and methods of analyzing a user's motion during a swimming session are described. One or more motions sensors can collect motion data of the user. A processor circuit can make motion analysis based on the motion data. The processor circuit can determine if the user's arm swing is a genuine swim stroke. The processor circuit can also determine whether the user is swimming or turning. The processor circuit can also classify the user's swim stroke style. The processor circuit can also determine the user's swim stroke phase. The processor circuit can also determine the user's stroke orbit consistency.

What is claimed is: 
     
         1 . A method for improving an accuracy of a wearable device while determining a user's arm motion, the method comprising:
 receiving, by a processor circuit of a wearable device, motion data from one or more motion sensors of the wearable device, wherein the one or more motion sensors comprises at least one of an accelerometer or a gyroscope;   determining, by the processor circuit using the motion data, rotational data expressed in a frame of reference based on the motion data;   calculating, by the processor circuit, a moment arm length based on the rotational data;   comparing, by the processor circuit, the moment arm length with a threshold length;   determining, by the processor circuit, the user's arm swing is a swim stroke based upon comparing the moment arm length with the threshold length;   calculating, by the processor circuit, at least one of a swimming metric or an energy expenditure of the user in response to determining the user's arm swing is a swim stroke, wherein the swimming metric comprises at least one of turns, breaths, laps, swim strokes, or swim stroke styles; and   outputting, by the processor circuit, the at least one of the swimming metric or the energy expenditure of the user.   
     
     
         2 . The method of  claim 1 , wherein the frame of reference is a body-fixed frame of reference with respect to the user device. 
     
     
         3 . The method of  claim 1 , wherein calculating the moment arm length comprises solving a least-squares equation. 
     
     
         4 . A method for improving an accuracy of a wearable device while determining a user is swimming, the method comprising:
 receiving, by a processor circuit of a wearable device, training data of the user;   receiving, by the processor circuit, motion data from one or more motion sensors of the wearable device, wherein the one or more motion sensors comprises at least one of an accelerometer or a gyroscope;   determining, by the processor circuit, a yaw angle of the user based on the motion data;   determining, by the processor circuit, a value of likelihood ratio corresponding to the yaw angle based on the training data;   comparing, by the processor circuit, the value of likelihood ratio with a threshold value;   determining, by the processor circuit, the user is swimming based upon comparing the value of likelihood ratio with the threshold value;   calculating, by the processor circuit, at least one of a swimming metric or an energy expenditure of the user in response to determining the user is swimming, wherein the swimming metric comprises at least one of turns, breaths, laps, swim strokes, or swim stroke styles; and   outputting, by the processor circuit, the at least one of the swimming metric or the energy expenditure of the user.   
     
     
         5 . A method for improving an accuracy of a wearable device while classifying a user's swim stroke style, the method comprising:
 receiving, by a processor circuit of a wearable device, motion data from one or more motion sensors of the wearable device, wherein the one or more motion sensors comprises at least one of an accelerometer or a gyroscope;   calculating, by the processor circuit, a fundamental period based on the received motion data;   determining, by the processor circuit, rotational data of the wearable device, wherein the rotational data is expressed in a frame of reference;   extracting, by the processor circuit, one or more features from the rotational data;   determining, by the processor circuit, the user's swim stroke style based on the one or more features; and   outputting, by the processor circuit, the determined swim stroke style.   
     
     
         6 . The method of  claim 5 , wherein the frame of reference is a body-fixed frame of reference with respect to the wearable device. 
     
     
         7 . The method of  claim 5 , wherein the frame of reference is an inertial frame of reference. 
     
     
         8 . The method of  claim 5 , wherein the one or more features comprise at least one of:
 a mean crown orientation of the wearable device,   a correlation of user's arm and wrist rotation, or   a contribution of rotation about a crown of the wearable device to a total angular velocity.   
     
     
         9 . The method of  claim 5 , wherein the one or more features comprise at least one of:
 a relative arm rotation about a band of the wearable device during a pull phase,   a moment arm of the user,   a ratio of acceleration z to rotation y,   a mean gravity crown weighted by acceleration,   a correlation between an orientation of top of a band of the wearable device and rotation around a band of the wearable device,   a root mean square (RMS) of a crown rotation,   a minimum rotation around a crown of the wearable device,   a maximum rotation around a band of the wearable device, or   a maximum rotation x over y.   
     
     
         10 . A method for improving an accuracy of a wearable device while determining phases of a user's swim stroke, the method comprising:
 receiving, by a processor circuit of a wearable device, motion data from one or more motion sensors of the wearable device, wherein the one or more motion sensors comprises at least one of an accelerometer or a gyroscope;   determining, by the processor circuit using the motion data, a first set of rotational data of the wearable device, wherein the first set of rotational data is expressed in a first frame of reference;   converting, by the processor circuit, the first set of rotational data of the wearable device into a second set of rotational expressed in a second frame of reference;   determining, by the processor circuit, a glide phase of a user's swim stroke based the second set of rotational data;   determining, by the processor circuit, a transition phase of the user's swim stroke based on the second set of rotational data;   determining, by the processor circuit, a pull phase and a recovery phase of the user's swim stroke based on the determined glide phase and transition phase;   calculating, by the processor circuit, one or more swimming metrics of the user based on the determined glide, transition, pull, and recovery phases of the user's swim stroke, wherein the one or more swimming metrics comprise at least one of turns, breaths, laps, swim strokes, or swim stroke styles; and   outputting, by the processor circuit, the calculated one or more swimming metrics of the user.   
     
     
         11 . The method of  claim 10 , wherein the first frame of reference is a body-fixed frame of reference with respect to the wearable device. 
     
     
         12 . The method of  claim 10 , wherein the second frame of reference is an inertial frame of reference. 
     
     
         13 . A method for improving an accuracy of a wearable device while determining a user's stroke orbit consistency, the method comprising:
 receiving, by a processor circuit of a wearable device, motion data from one or more motion sensors of the wearable device, wherein the one or more motion sensors comprises at least one of an accelerometer or a gyroscope;   determining, by the processor circuit using the motion data, rotational data of the user device, wherein the rotational data is expressed in a frame of reference;   determining, by the processor circuit, a first direction along which the rotational data have the least variance in a first past period;   determining, by the processor circuit, a second direction along which the rotational data have the least variance in a second past period;   determining, by the processor circuit, a difference between the first direction and the second direction;   determining, by the processor circuit, a stroke orbit consistency of the user based on the difference between the first direction and the second direction; and   outputting, by the processor circuit, the determined stroke orbit consistency of the user.   
     
     
         14 . The method of  claim 13 , wherein the frame of reference is a body-fixed frame of reference with respect to the user device. 
     
     
         15 . The method of  claim 13 , wherein the frame of reference is an inertial frame of reference. 
     
     
         16 . The method of  claim 13 , wherein determining the first direction and the second directing comprises determining an axis of rotation. 
     
     
         17 . The method of  claim 13 , wherein the first past period is substantially 10 seconds. 
     
     
         18 . The method of  claim 13 , wherein the second past period is substantially 3 minutes. 
     
     
         19 . A method for improving an accuracy of a wearable device while determining a user's stroke orbit consistency, the method comprising:
 receiving, by a processor circuit of a wearable device, motion data from one or more motion sensors of the wearable device, wherein the one or more motion sensors comprises at least one of an accelerometer or a gyroscope;   determining, by the processor circuit using the motion data, rotational data expressed in a frame of reference;   determining, by the processor circuit, a histogram of the user's stroke orbit using the rotational data;   determining, by the processor, a level of entropy based on the histogram;   determining, by the processor circuit, a level of orbit consistency of the user based on the determined level of entropy; and   outputting, by the processor circuit, the determined level of orbit consistency of the user.   
     
     
         20 . The method of  claim 19 , wherein the frame of reference is a body-fixed frame of reference with respect to the user device. 
     
     
         21 . The method of  claim 19 , wherein the frame of reference is an inertial frame of reference.