Automatic fall detection based on Doppler radar motion signature

Falling is a common health problem for elderly. It is reported that more than one third of adults 65 and older fall each year in the United States. To address the problem, we are currently developing a Doppler radar-based fall detection system. Doppler radar sensors provide an inexpensive way to recognize human activity. In this paper, we employed mel-frequency cepstral coefficients (MFCC) to represent the Doppler signatures of various human activities such as walking, bending down, falling, etc. Then we used two different classifiers, SVM and kNN, to automatically detect falls based on the extracted MFCC features. We obtained encouraging classification results on a pilot dataset that contained 109 falls and 341 non-fall human activities.