| In the 21 st century,the elderly in the world is increasing rapidly and the aged tendency of population is becoming evident.The changes in the distribution of the population structure have a far-reaching influence on all aspects of the society.Now as lifestyle changes,there are more empty nesters.The lack of corresponding public provision for the aged can't meet the need of the elderly.Some studies show that fall has become an important factor threatening health of the elderly,but we can't watch them all day long.With the popularity of Android device,we want to develop a fall detection application.And it will protect the elderly from being hurt from falls.This thesis implement a fall detection application based on the built in 3-axis accelerometer and gyroscope on the Android device.First,the thesis studies the the characteristic of different daily activities according to the measured sensor data.Then,the thesis discusses support vector machine(SVM)and the algorithm of accelerometer combined with gyroscope.This thesis studies SVM classification algorithm,discusses the classifier and the classification model we used.This thesis accomplished verification procedures on the PC side,and uses LIBSVM for data training and classification.According to the analysis of the classification results,we find the SVM algorithm is more suitable as a pre-process method and choose the algorithm of accelerometer combined with gyroscope for implemention in the project.We present the components we used in Android development,select the corresponding characteristics,design the algorithm of accelerometer combined with gyroscope to detect falls and use a warning dialog to reduce the rate of false positives.Finally,we implement this algorithm by using Android Studio development environment and test the fall detection application we accomplished,the test result indicates that the implemention of the algorithm of accelerometer combined with gyroscope is verified in accordance with the design. |
PDF Full Text Request