Abstract
In this paper, a wearable device designed using acceleration sensors to obtain acceleration information generated by the user's muscles during exercise follows data collection, data pre-processing, feature extraction and feature selection, classification model training, and evaluation of this muscle information of human movement. Designed a relevant model verification system and wearable device, and carried out verification experiments on the model in real-time, accuracy, interactivity, and parallelism with other sensor data types. Transitional motion detection and segmentation algorithms can effectively segment out the transition actions included in the acceleration sequence, and using this method to classify and recognize nine kinds of human motion information, the average recognition rate reaches 98.56%.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call