Human Motion Data Detection and Classification

Abstract

Falling down can be a major reason for injuries and it is a typical problem for elderlies, especially for those who live alone. There are more death incidents that is caused by falling down where no one notices until someone comes across the incident without any notice. Thus a system is required to automatically detect when an elderly has fallen down, and alert the caregivers or others of the incident in order to provide immediate assistance. This project contains two parts, study of fall and develop a prototype to detect fall. The first one includes the data collection and data analysis on fall event. The second part present a solution to keep track of patient movements, classify the fall event and alert their care givers through the cloud system, which will include the whole process of development and limitations of the development.