Activity Recognition Technology Based On Smart Phone For Behavioral Monitoring Of The Elderly

Now,smart phones are equipped with a large number of sensors,computing and storage resources,this makes the various sensor based applications appear in various fields,such as personal health,environmental monitoring,social networking.At present,the smart phones integrate a wealth of sensors,such as accelerometers,gyroscopes,GPS,microphone,camera,luminance sensor,Wi-Fi and Bluetooth interface,provide a convenient platform for human activity recognition.In addition,it also provides a more simple sensor data acquisition for researchers.Based on the combination of GPS module and the accelerometer embedded in the smart phone,this thesis proposed a method using acceleration data and GPS velocity data to realize real-time indoor and outdoor activity recognition for behavioral monitoring of the elderly.The main work of this thesis is as follows:(1)The workflow and research methods of the behavior recognition based on the accelerometer are summarized.This thesis proposed a method of activity recognition technology based on smart phone for behavioral monitoring of the elderly,based on the accelerometer and GPS module in smart phone,a new behavioral data model is proposed.(2)This thesis study the segmentation of the sensor data,and select the appropriate size sliding window to segment the human behavioral data.Finally,the results of corresponding contrast experiments show that the accuracy of the classification model which uses eigenvalue is higher than that doesn't uses.In particular,the recognition accuracy of human activity by car has been greatly improved.(3)On the basis of the data model and activity recognition method based on the smart phone acceleration sensor and GPS module,this thesis designs and implements the real-time recognition system of human motion patterns.Firstly,the system realizes the real-time reception of the human behavior data from the mobile terminal and the background processing and classification;Finally,the user activity patterns are classified as 6 types: sitting,standing,walking,running,cycling and riding.