Font Size: a A A

Design And Implementation Of Human Fall Detection And Alarm System Based On Smart Phone

Posted on:2017-01-11Degree:MasterType:Thesis
Country:ChinaCandidate:G P ZhangFull Text:PDF
GTID:2308330503469339Subject:Computer technology
Abstract/Summary:PDF Full Text Request
Nowdays,the speed of social aging is developing more quickly in our country. By 2030, the population of the elder are nearly 300 million. Fall is an important factor that affect the eldery’s daily physicial and health in china. The occurrence of eldery’s fall can be real-time detected and send a distress signal to minimize injury to the body caused by a fall has a great significance. Android smart phone platform is free and open soure for mobile devices. Today, the functions of mobile phone is diversified. Mobile phones can be used in medical aid. This thesis analyzes the characteristics of human falling-action and researchs the human fall detction and alarm algorithm,and then designs and implements a human fall detction and alarm system based on android smart phone platform.Firstly, we analyze the technologies of data acquisition and preprocessing based on Android platform. We study the feature selection method of sensor’s data and the classification algorithm of human gait.Secondly, we research and design a human fall detection algorithm. We search the data’s feature by using the training data and classify the training data by using classifiers of J48,Bayes and SMO. According to the correct rate of classification results,we use J48 classification algorithm and decision tree as the basis algorithm of the fall detection algorithm in this paper. The correct rate of fall detection is 92% by using the training data to test fall detection algorithm, so this algorithm is an effective algorithm.Thirdly, we design the architecture of human fall detction and alarm system. It is composed of the sensor driver, the human fall detection alarm service and the user application. We implement the hardware driver of BMA250 sensor. It is registered as an 12 C client and a platform device.Finally, we implement human fall detection and alarm system and applications. We design and implement data communication between the sensor service and human fall detection service, and finish the switch function of the system. When a fall event is detected,the service will start an alarm for helping.
Keywords/Search Tags:triaxial acceleration sensor, fall detection, J48 classifier, Android platform
PDF Full Text Request
Related items