Contactless Fall Detection using Doppler Radar

Abstract

Falling is the main cause of disability and the fatality of elderly. In this work, a 24 GHz continuous wave (CW) Doppler Radar-based novel system is proposed as a cheap, easy to use and effective solution for elderly fall detection. A set of features extracted from both time and frequency domains is examined and features that contribute most to distinguish between fall and non-fall samples are selected. Finally, different learning techniques including support vector machine (SVM), k nearest neighborhood (kNN), Naive Bayes (NB), linear discriminant analysis (LDA)and decision tree (DT) are evaluated and linear discriminant analysis method is selected as the most accurate classification model. The proposed system performed fall detection with 88% accuracy.