Research Of Smart Garments Oriented Multi-Physiology Information Fusion For Mood Distinction

As a smart health monitor system,the smart clothing bears the function to detect a range of physiological signals from the human body.By using the medical diagnostic platform,in order to achieve the health care goal,it uses the acquisition physiological signals for online health analysis,diagnosis and treatment.Due to the various organs of the body work coordinately and interact with each other,a single physical indicator can not adequately reflect the body's state of health.In recent year,the medical information of multi-mode technology gets much attention on clinical,it monitors two or more physiological signals and combines their advantages.It can make up for a doctor to determine a comprehensive artificial reasoning and subjective and improves the efficiency and reliability of medical diagnosis.The main contributions are as follows:Firstly,this thesis first introduces some background on wearable intelligent system. Based,on the basis of smart clothing on-line medical diagnosis,we proposed to the establishment of Smart Garments Multi-physiology information fusion for mood distinction.Secondly,through research on the multi-source information fusion theory,we discuss the fusion theory of classification algorithms and the applications,focused on BAYES discrimination,BP neural networks,fuzzy set theory and the D-S evidence theory.We also introduce SVM and its theoretical basis-statistical learning theory,the SVM core idea:kernel funcation,as well as SVM of the dual classification algorithms and multi-classification algorithm.And,we compare these four algorithms and SVM of the advantages and disadvantages.Next,from the physical point of view,the thesis introduces three physiological signals for model-building,such as electrocardiogram(ECG),respiration(RSP),body temperature(TEM),and we regard that the introduction of RSA RSA(Respiratory Sinus Arrhythmia) physiological phenomena is good for the emotional recognition rate.By studying the three types of physiological signal acquisition and processing methods,and the multi-physiology information fusion methods and applications,we focuse on studying SVM as a classifying and fusion algorithm in the multi-physiology information fusion application.Then,from the perspective of physiological mechanism,the thesis introduces three physiological signals for the model establishment.We also study the feature extraction method,and introduce sequence minimization method because the selection of characteristics is good or bad impact on the recognition rate directly.And through the feature of physiological signals,we study the multi-physiology information fusion method and its application.By acquisiting the three-induced physiological signals in a particular environment,combined with the subjective mood questionnaire,we do feature extraction and classification on the three physiological signal by using MATLAB language based on SVM and establish the Multi-physiology Information Fusion Mood Distinction to obtain a high rate of emotional identification.Finnally,through LABVIEW and MATLAB,we implements the Smart Garments Multi-physiology Information Fusion Platform for Mood Distinction.It provides a harmonious environment for human-computer interaction.