Design And Realization Of Fall Detection System Based On Inertial Sensing

Population aging is a worldwide problem.According to the World Health Organization definition of population aging and China's sixth national census data,China has entered aging society.In daily life,the elderly were prone to fall and other unexpected events will occur.CCTV has reported that in China about 40 million people aged over 65 accidentally fell each year.When the elderly fall,it will cause irreversible consequences without timely aid.So the fall detection technology has a great practical value.The specific work of this paper is:1)Carry out an experiment to find out the impact of inertial sensor node location on experimental data and select the waist as a wear position.Compare the samples in real life with the samples obtained by the experiment and analyze the characteristics of two samples.Analyze the requirements and set the design goal of the system.On this basis,the specific technical scheme of the system is putted forward.2)Collect the inertia data samples of the fall events and daily behavior activities,and build data sets for the algorithm design.The pre-processing of the original data set is carried out.The Relief algorithm is used to select the feature,selecting the feature set for the classifier.Take the system development into account,compare the specificity and sensitivity of the algorithm and select the decision tree as the classification model.3)Achieve a system of the cloud server,Android applications and inertial sensor nodes.When it detects a fall event,the inertial sensor node sends the information to the smartphone.The phone locates the user and sends help information to the cloud server,while the cloud server informs the community or hospital personnel to aid the user.4)Verify the function of the entire system and the results suggest that it can achieve all functions correctly.The sensitivity and specificity of the fall detection algorithm are tested.The test results show that the sensitivity of the fall detection algorithm is 95.56% and the specificity is 98.00%.It can detect the fall accurately.Through the analysis and comparison of the current situation of domestic and foreign research on fall detection,this paper chooses the detection scheme based on inertial sensing signal and combines with the cloud server technology and the Andrews smart phone to implement a set of fall detection system based on inertial sensing.It is fully functional and has a good practical value.