Leveraging machine learning and fractal analysis for classifying motion

Abstract

A machine learning and fractal analysis process for classifying human or animal motion, including the classification of patterns generated by human or animal motion in order to assess the quality of athletic performance, artistic performance, form, or other quality of motion.

What is claimed is: 
     
         1 . A method, comprising:
 sensing sensor data generated by one or more inertial sensors, the sensor data being caused by movement of a user, and the sensor data comprising acceleration data in multiple dimensions;   recording the sensor data with a wearable computing system configured to be worn by the user;   projecting, by the wearable computing system, a three-dimensional acceleration pattern indicated by the acceleration data to a two-dimensional plane to generate at least one two-dimensional acceleration projection having a first acceleration dimension and a second acceleration dimension, where the first acceleration dimension is orthogonal to the second acceleration dimension and both the first and second acceleration dimensions have units of distance per second squared;   providing to the user, by the wearable computing system, visual feedback regarding the user movement based on the at least one two-dimensional acceleration projection;   wherein the visual feedback comprises at least one two-dimensional orbital image showing an orbital pattern of the user movement in the first and second acceleration dimensions.   
     
     
         2 . The method of  claim 1 , wherein the at least one two-dimensional orbital image comprises a series of data points computed based on the sensor data, and a curve joining the series of data points. 
     
     
         3 . The method of  claim 2 , wherein the at least one two-dimensional orbital image comprises a plurality of orbital cycles and the user movement comprises a plurality of repetitive user movement cycles; each orbital cycle of the plurality of orbital cycles representing an individual movement cycle of the plurality of repetitive user movement cycles. 
     
     
         4 . The method of  claim 3 , wherein each orbital cycle represents an individual stride of the user. 
     
     
         5 . The method of  claim 3 , wherein the at least one two-dimensional orbital image comprises a plurality of pixels, each pixel having an associated pixel value; and
 wherein for each individual pixel of the plurality of pixels, the associated pixel value is increased each time an individual orbital cycle of the plurality of orbital cycles overlaps the individual pixel.   
     
     
         6 . The method of  claim 1 , wherein projecting the three-dimensional acceleration pattern to the at least one two-dimensional plane comprises projecting the three-dimensional acceleration pattern to multiple two-dimensional planes to generate corresponding two-dimensional acceleration projections that each have a first dimension that has units of distance per second squared in a first direction and a second dimension that has units of distance per second squared in a second direction, where the first direction is orthogonal to the second direction. 
     
     
         7 . The method of  claim 6 , wherein the multiple two-dimensional planes comprise a side plane, a front plane, and a top plane. 
     
     
         8 . The method of  claim 1 , further comprising analyzing the at least one two-dimensional orbital image to determine whether it is indicative of potential injury or inefficient motion. 
     
     
         9 . The method of  claim 1 , further comprising analyzing the at least one two-dimensional orbital image to determine whether it corresponds with a proper movement form. 
     
     
         10 . The method of  claim 1 , further comprising transmitting the at least one two-dimensional acceleration projection to a remote computing device for analysis. 
     
     
         11 . A wearable computing device configured to be worn by a user, the wearable computing device comprising:
 one or more inertial sensors configured to generate sensor data based on movement of the user, the sensor data comprising acceleration data in multiple dimensions;   at least one processor configured to:
 receive the sensor data generated by the one or more inertial sensors; 
 project a three-dimensional acceleration pattern indicated by the acceleration data to a two-dimensional plane to generate at least one two-dimensional acceleration projection that has a first acceleration dimension and a second acceleration dimension, where the first acceleration dimension is orthogonal to the second acceleration dimension and both the first and second acceleration dimensions have units of distance per second squared; 
 provide visual feedback of the user movement based on the at least one two-dimensional acceleration projection; 
   wherein the visual feedback comprises at least one two-dimensional orbital image showing an orbital pattern of the user movement in the first and second acceleration dimensions.   
     
     
         12 . The wearable computing device of  claim 11 , wherein the at least one two-dimensional orbital image comprises a series of data points computed based on the sensor data, and a curve joining the series of data points. 
     
     
         13 . The wearable computing device of  claim 12 , wherein the at least one two-dimensional orbital image comprises a plurality of orbital cycles and the user movement comprises a plurality of repetitive user movement cycles; where each orbital cycle of the plurality of orbital cycles represents an individual movement cycle of the plurality of repetitive user movement cycles. 
     
     
         14 . The wearable computing device of  claim 13 , wherein each orbital cycle represents an individual stride of the user. 
     
     
         15 . The wearable computing device of  claim 13 , wherein the at least one two-dimensional orbital image comprises a plurality of pixels, each pixel of the plurality of pixels having an associated pixel value; and
 wherein for each individual pixel of the plurality of pixels, the associated pixel value is increased each time an individual orbital cycle of the plurality of orbital cycles overlaps the individual pixel.   
     
     
         16 . The wearable computing device of  claim 11 , wherein the at least one processor is further configured to analyze the at least one two-dimensional orbital image to determine whether it is indicative of potential injury or inefficient motion. 
     
     
         17 . The wearable computing device of  claim 11 , wherein the at least one processor is further configured to analyze the at least one two-dimensional orbital image to determine whether it corresponds with a proper movement form. 
     
     
         18 . The wearable computing device of  claim 1 , further comprising a transmitter configured to transmit the at least one two-dimensional acceleration projection to a remote computing device for analysis. 
     
     
         19 . A method, comprising:
 sensing sensor data generated by one or more inertial sensors, the sensor data being caused by movement of a user, and the sensor data comprising acceleration data in multiple dimensions;   recording the sensor data with a wearable computing system configured to be worn by the user;   projecting, by the wearable computing system, a three-dimensional acceleration pattern indicated by the acceleration data to a two-dimensional plane to generate at least one two-dimensional acceleration projection having a first acceleration dimension and a second acceleration dimension, where the first acceleration dimension is orthogonal to the second acceleration dimension and both the first and second acceleration dimensions have units of distance per second squared;   analyzing the at least one two-dimensional acceleration projection using fractal analysis at multiple scales and taking into account variations in fractal dimensions at different scales:   providing to the user, by the wearable computing system, visual feedback regarding the user movement based on the at least one two-dimensional acceleration projection;   wherein the visual feedback comprises at least one two-dimensional orbital image showing an orbital pattern of the user movement in the first and second acceleration dimensions.   
     
     
         20 . The method of  claim 19 , wherein the at least one two-dimensional orbital image comprises a plurality of orbital cycles and the user movement comprises a plurality of repetitive user movement cycles; each orbital cycle of the plurality of orbital cycles representing an individual movement cycle of the plurality of repetitive user movement cycles.