Recognizing Human Activity in Free-Living Using Multiple Body-Worn Accelerometers

Abstract

Recognizing human activity is very useful for an investigator about a patient's behavior and can aid in prescribing activity in future recommendations. The use of body worn accelerometers has been demonstrated to be an accurate measure of human activity; however, research looking at the use of multiple body worn accelerometers in a free living environment to recognize a wide range of activities is not evident. This paper aimed to successfully recognize activity and sub-category activity types through the use of multiple body worn accelerometers in a free-living environment. Ten participants (Age = 23.1 +/- 1.7 years, height = 171.0 +/- 4.7 cm, and mass = 78.2 +/- 12.5 Kg) wore nine body-worn accelerometers for a day of free living. Activity type was identified through the use of a wearable camera, and subcategory activities were quantified through a combination of free-living and controlled testing. A variety of machine learning techniques consisting of preprocessing algorithms, feature, and classifier selections were tested, accuracy, and computing time were reported. A fine k-nearest neighbor classifier with mean and standard deviation features of unfiltered data reported a recognition accuracy of 97.6%. Controlled and free-living testing provided highly accurate recognition for sub-category activities (> 95.0%). Decision tree classifiers and maximum features demonstrated to have the lowest computing time. Results show that recognition of activity and sub-category activity types is possible in a free-living environment through the use of multiple body worn accelerometers. This method can aid in prescribing recommendations for activity and sedentary periods for healthy living.