Design Of Fall Detection Software Based On Android Platform

As the aging process intensifying in C hina, the increasing number of the Chinese empty nester calls people’s attention to the elder’s healthy condition. The elder’s fall has the features of high probability, high damage and low salvation rate, and usually causes bad results like Coma, paralysis even death. Therefore, it’s necessary to figure out a portable real-time system, which can give an accurate judgment whether the elder falling down, through that system, much time can be saved for the first aid. This project has a significant in healthy guarantee on elder and improving the level of medical care.Recently, there are a large number of domestic and foreign research about the fall detection, and the mature products have already been put into use, but they do not have historical data storage function of the fall, the scope of application about fall detection in social networks is limited, the tightness between product quality of service and user is not good enough, and users’ comfort level is not so good.This paper designs a system for the elder’s fall detection, the detection accuracy has been improved, and the function like wireless location, automatic alarm and remote alarm can make the injured get the rescue and treatment in time. This paper mainly includes three parts:① In order to improve the accuracy of the fall detection, under the consideration of the practicability and accuracy, this paper chooses the support vector machine as the algorithm of the human fall detection.② Elastic Compute Service platform function design: According to the cloud server in data storage with high reliability, general characteristics, the cloud server Sina SAE is taken as the fall detection system’s storage center, which has the function of data storage, statistical analysis, information push etc.③ Design of mobile software based on Software Engineering: According to the software engineering theory and the present development situation, based on the Android system this paper designs the fall detection system software in the background, which can receive real-time alarm information from the fall detector. The relatives can directly call the map software in the mobile phone to query the location information of the fall detector, which facilitates tracking and monitoring.According to the fall detection experiment design, the result analysis verifies the algorithm of supporting vector machine can achieve a high detective accuracy as expected. The system in this paper can not only achieve a high detection rate, reduce the false alarm rate, but also combine with the mobile phone terminal software and cloud server at the same time, which can achieve a series of process such as judging alarm information, analyzing and sending the warning information, etc.