Research On Human Movement Information Collection And Classification Technology Based On Smart Bracelet

The technology of the human motion information recognition has universal application prospects and has potential application value in the areas of human behavior monitoring,medical monitoring,medical rehabilitation,interactive games,virtual sports scenes,film special effects and interactive learning.This article designs a set of identification system that collects information on the mobile devices.It achieves the real-time recognition of human motion through the data analysis of the smart wristband acceleration sensor.This article is based on the smart bracelet worn daily,using the built-in three-axis accelerometer data,using the support vector machine and the convolutional neural network to analyze the wrist movement stance.After comparison,the convolutional neural network has a better recognition effect.The use of convolutional neural networks for human gesture recognition is the innovations in this article.For these contents,the main work of this article is as follows:1.A human body information collection system based on a smart wristband was designed and related program software was developed.2.The recognition algorithm is constructed by using support vector machines in machine learning.This method requires feature extraction and feature selection of sample data to obtain key sample information.This algorithm can achieve up to 94%recognition accuracy.3.This article designed a recognition algorithm using convolutional neural networks based on deep learning technology.This method does not require separate execution of feature extraction and feature recognition steps,and the process implicitly integrates into the classification algorithm.This paper compares the parameters of convolutional neural network and obtains the optimal network structure model.The network model is then transfered to the mobile and successfully implemented.In general,the collection and classification technology of human motion information based on smart wristbands is still in the development stage.The research content of this topic has practical application requirements and theoretical research value,and has in-depth research significance and value.