Automated motion data processing

Abstract

Motion data processing includes receiving a plurality of motion data samples, wherein the plurality of motion data samples comprises a plurality of acceleration data samples obtained by an accelerometer and a plurality of angular velocity data samples obtained by a gyroscope, and the plurality of motion data samples pertains to a plurality of motions performed by a user; and correcting for distortions in the motion data samples, including determining, based at least in part on some of the motion data samples, a set of one or more corrected core measurements of the user's body positions during a subset of the plurality of motions.

What is claimed is: 
     
         1 . A motion data processing system, comprising:
 one or more processors configured to:
 receive a plurality of motion data samples pertaining to a plurality of motions performed by a user; 
 determine, based at least in part on the plurality of motion data samples, a set of one or more corrected core measurements corresponding to one or more body positions of the user during the plurality of motions, the set of one or more corrected core measurements comprising interpolated orientation data based at least in part on an interpolation of corrected orientation data and current orientation data obtained from the plurality of motion samples; and 
 provide feedback to the user, the feedback being based at least in part on a result of the comparing the one or more corrected core measurements to the set of reference motions; and 
   one or more memories coupled to the one or more processors, configured to provide the one or more processors with instructions.   
     
     
         2 . The system of  claim 1 , wherein the plurality of motion data samples further comprises a plurality of magnetic field data samples. 
     
     
         3 . The system of  claim 1 , wherein the set of one or more corrected core measurements is determined based at least in part on a statistical model. 
     
     
         4 . The system of  claim 1 , wherein to determine the current orientation data sample includes performing a quaternion transform of a previous interpolated and corrected orientation data sample and a current angular velocity data sample. 
     
     
         5 . The system of  claim 1 , wherein the plurality of data samples comprise a plurality of acceleration data samples obtained by an accelerometer and a plurality of angular velocity data samples obtained by a gyroscope. 
     
     
         6 . The system of  claim 5 , wherein the plurality of acceleration data samples and the plurality of angular velocity data samples are preprocessed, the preprocessing including to detect a lost data sample and to estimate the lost data sample by interpolation of received data samples. 
     
     
         7 . The system of  claim 1 , wherein the set of one or more corrected core measurements further comprises corrected velocity data that is based at least in part on orientation data and a statistical model. 
     
     
         8 . The system of  claim 7 , wherein the statistical model is a distribution model of sensor output features to ground truth motion velocity. 
     
     
         9 . The system of  claim 7 , wherein the orientation data used to determine the velocity data is the interpolated orientation data. 
     
     
         10 . The system of  claim 7 , wherein to determine the corrected velocity data includes to:
 determine an acceleration in world space;   integrate the acceleration to determine a current velocity data sample;   determine a probability distribution of a ground truth velocity based at least in part on the statistical model;   determine an offset velocity to maximize a sum of probabilities for a sample window; and   determine a corrected current output velocity based at least in part on the probability distribution and the offset velocity.   
     
     
         11 . The system of  claim 1 , wherein the one or more processors are further to determine a form measurement that characterizes a form of the user during a subset of the plurality of motions. 
     
     
         12 . The system of  claim 11 , wherein the form measurement comprises a measurement pertaining to one or more of cadence, stride, range of motion, efficiency, speed, landing impact, and pronation. 
     
     
         13 . The system of  claim 11 , wherein to determine the form measurement includes to:
 determine a dominant rotational axis based at least in part on the set of the corrected core measurements; and   identify a first set of samples involved in a first rotational direction and a second set of samples that are involved in a second rotational direction based on a projection on the dominant rotational axis.   
     
     
         14 . The system of  claim 11 , wherein to determine the form measurement further includes to identify distinct parts of the set of corrected core measurements based on pattern recognition; and
 determine the form measurement based on the identified distinct parts.   
     
     
         15 . The system of  claim 1 , wherein the set of one or more corrected core measurements includes a quaternion vector associated with a subset of the plurality motions and a velocity vector associated with the subset of the plurality motions. 
     
     
         16 . The system of  claim 1 , wherein the feedback to the user is provided on user interface. 
     
     
         17 . The system of  claim 16 , wherein the feedback comprises a recommendation for the user to adjust a form of an exercise. 
     
     
         18 . The system of  claim 16 , wherein the feedback comprises a difference between the plurality of motions performed by a user and the set of reference motions. 
     
     
         19 . A method comprising:
 receiving, by one or more processors, a plurality of motion data samples pertaining to a plurality of motions performed by a user;   determining, based at least in part on the plurality of motion data samples, a set of one or more corrected core measurements corresponding to one or more body positions of the user during the plurality of motions, the set of one or more corrected core measurements comprising interpolated orientation data based at least in part on an interpolation of corrected orientation data and current orientation data obtained from the plurality of motion samples; and   providing, by the one or more processors, feedback to the user, the feedback being based at least in part on a result of the comparing the one or more corrected core measurements to the set of reference motions.   
     
     
         20 . A computer program product for motion data processing, the computer program product being embodied in a tangible computer readable storage medium and comprising computer instructions for:
 receiving, by one or more processors, a plurality of motion data samples pertaining to a plurality of motions performed by a user;   determining, based at least in part on the plurality of motion data samples, a set of one or more corrected core measurements corresponding to one or more body positions of the user during the plurality of motions, the set of one or more corrected core measurements comprising interpolated orientation data based at least in part on an interpolation of corrected orientation data and current orientation data obtained from the plurality of motion samples; and   providing, by the one or more processors, feedback to the user, the feedback being based at least in part on a result of the comparing the one or more corrected core measurements to the set of reference motions.