An effective video-based model for fall monitoring of the elderly

Abstract

Fall is one of the major health challenges facing the elderly adults, especially the adults with high fall risk factors. In this paper, we aim to build a video-based model to mitigate the consequences of fall of the elderly at two application scenarios: (1) predict the fall risk caused by unbalanced gait and (2) detect a fall event as soon as it happens. In the first stage, we use a common camera to capture the video of a person moving such as doing actions or walking. In the second stage, our proposed model for both scenarios follows the same three-module structure: (1) human object segmentation using background subtraction, (2) feature representation using two separate feature descriptors, one for each application scenario, and (3) abnormal event detection based on Hidden Markov Model. The final stage is to convey an SMS message to the pre-defined cell phone number to notify the caregiver of detected anomaly. Experimental results show the promising performance of the proposed model in terms of the acceptable accuracy and the low processing time.