Elderly Health Monitoring System with Fall Detection Using Multi-Feature Based Person Tracking

Abstract

The need for personalized surveillance systems for elderly health care has risen drastically. However, recent methods involving the usage of wearable devices for activity monitoring offer limited solutions. To address this issue, we have proposed a system that incorporates a vision-based deep learning solution for elderly surveillance. This system primarily consists of a novel multi-feature-based person tracker (MFPT), supported by an efficient vision-based person fall detector (VPFD). The MFPT encompasses a combination of appearance and motion similarity in order to perform effective target association for object tracking. The similarity computations are carried out through Siamese convolutional neural networks (CNNs) and long-short term memory (LSTM). The VPFD employs histogram-of-oriented-gradients (HoGs) for feature extraction, followed by the LSTM network for fall classification. The cloud-based storage and retrieval of objects is employed allowing the two models to work in a distributed manner. The proposed system meets the objectives of ITU Focus Group on AI for Health (FG-AI4H)under the category, “falls among the elderly”. The system also complies with ITU-T F.743.1 standard, and it has been evaluated over benchmarked object tracking and fall detection datasets. The evaluation results show that our system achieves the tracking precision of 94.67% and the accuracy of 98.01% in fall detection, making it practical for health care system use. The HoG feature-based LSTM model is a promising item to be standardized in ITU for fall detection in elderly healthcare management under the requirements and service description provided by ITU-T F.743.1.