Design Of Fall Detection System For Elderly Based On MPU6050

With the development of the modern medical technology people pay more attention on the heath of themselves in their daily life,human life has been greatly extended,but at the same time,it also brings the problem of aging population.With the increasing problem of aging,how to ensure the normal life of the elderly and their safety,is the social focus of concern..Among of them,fall detection is a major challenge for the Elderly Security problems.The problem of falls in elderly people will lead to higher rates of hospitalization,medical treatment and mortality,as a burden to the family and society.Nowadays,wearable fall detection equipment has received much attention in academia and industry.Therefore,the research of the elderly fall detection equipment is of great significance.In this paper,a fall detection system is designed for the fall problem of the elderly.This paper takes into account the portability and real-time performance of the fall detection equipment,and designs a wearable old people fall detection device based on MPU6050.The main work of this paper is as follows:(1)The design of human fall detection system.Firstly,we analyze the movement state of the human body during the fall,so as to select the best feature extraction part of the human body during the fall process;and then then set up the system of the human fall detection,including the hardware composition of the system,the software architecture and the system parameter settings,etc.,which mainly uses the system MPU6050 as the main controller.(2)The method of extracting and classifying the characteristics of human fall process.Firstly,the system collects and preprocesses the samples of the human body's daily behavior and the human body's falling data.The purpose of the preprocessing is to eliminate the effect of noise in the data.Then,the fast Fourier Transform(FFT)of the preprocessed data is used and the Principal Component Analysis(PCA)is used to reduce the dimension of frequency domain data to obtain the eigenvalues of human body's daily behavior and human falls.Finally,Extreme Learning Machine(ELM)classification method is used to establish the body fall recognition model.(3)Experiment and result analysis.The experiment process is designed and identified by the Extreme Learning Machine and the Support Vector Machine(SVM).The experimental results show that the feature extraction and classification method proposed in this paper is suitable for human fall detection,and the system of the fall detection has good effect.