Exercise Evaluation Using Wrist Sensors
Naveen Manoharan, Kewei Sha, Jiang Lu · 2017
Exercises are effective and efficient for both major disease recovery and prevention. Home-based exercises significantly reduce the cost of the exercise, the quality of the home-based exercises, however, needs to be monitored and evaluated to guarantee the effectiveness of the exercises. This work proposes a lightweight algorithm to evaluate the accuracy of the performed exercise by analyzing the data collected from sensors attached to the wrist of the exercisers. When the exercise is performed at home, features that model exercise are extracted from the collected sensor data. Then these features are compared with a pre-stored true model defined by exercise coaches, and exercise accuracy evaluation is generated.