Matching System for Correlating Accelerometer Data to Known Movements

Abstract

The present invention extends to methods, systems, and computer program products for providing a matching system for correlating accelerometer data to known movements. Data representing known movements can be obtained and stored in a database such as by processing and storing accelerometer data obtained from one or more accelerometers worn by a user while performing a particular movement. The accelerometer data obtained from a particular movement can be processed to generate a feature set descriptive of the accelerations associated with a particular movement or series of movements.

What is claimed: 
     
         1 . A method for identifying a particular movement from accelerometer data by comparing an identified sequence in the accelerometer data to known sequences, the method comprising:
 storing a plurality of entries in a database on a portable computing device, each entry representing one or more known sequences of accelerometer data that are generated when a particular movement is performed;   receiving accelerometer data from one or more accelerometers worn by a user while performing a first movement;   accessing the database to determine that the accelerometer data received from the one or more accelerometers includes the one or more known sequences of a first entry; and   determining that the first entry is associated with a first particular movement.   
     
     
         2 . The method of  claim 1 , further comprising:
 displaying an indication that the first particular movement has been performed by the user.   
     
     
         3 . The method of  claim 2 , wherein displaying the indication comprises incrementing a count of the number of times the user has performed the first particular movement. 
     
     
         4 . The method of  claim 1 , wherein at least some of the entries in the database include known sequences generated by more than one accelerometer when the corresponding particular movement is performed. 
     
     
         5 . The method of  claim 1 , further comprising:
 creating a new entry in the database, the new entry containing sequences of accelerometer data received from one or more accelerometers while the user is performing a custom movement.   
     
     
         6 . The method of  claim 1 , wherein each entry in the database comprises a feature set, and
 wherein accessing the database to determine that the accelerometer data received from the one or more accelerometers includes the one or more known sequences of a first entry further comprises converting the received accelerometer data into a feature set of an unknown movement.   
     
     
         7 . The method of  claim 6 , wherein converting the received accelerometer data into a feature set of an unknown movement comprises:
 extracting a chunk of the accelerometer data, the chunk comprising accelerometer data received over a particular interval of time, the accelerometer data including accelerometer data corresponding to a plurality of axes and one or more locations;   splitting the chunk into a plurality of time series, each time series including the accelerometer data from a particular axis;   for each time series, extracting the magnitude of a plurality of frequencies in each time series;   for each time series, creating a group of summed magnitudes, each group comprising the sum of the magnitudes of a subset of the plurality of frequencies; and   creating the feature set by aggregating the summed magnitudes of each subset of each time series into a list.   
     
     
         8 . The method of  claim 6 , wherein determining that the accelerometer data received from the one or more accelerometers includes the one or more known sequences of a first entry comprises:
 for each of a plurality of feature sets in the database, determining the inverse Euclidean metric of the feature set of the unknown movement and the feature set; and   determining which feature set of the plurality of feature sets in the database yields the greatest inverse Euclidean metric.   
     
     
         9 . The method of  claim 1 , wherein the entries in the database include a speed factor that identifies the relative speed at which the associated particular movement is performed.