Person Fall Recognition by using Deep Learning: Convolutional Neural Networks and Image category classification using bag of feature
Faria Ferooz, Muhammad Adeel Ashraf, Waqar Hussain, Ahmad Hassan Butt, Yaser Daanial Khan · 2019
The rate aging people living alone at home are increasing day by day. Fall is one of the major risks for elderly people. Most of the elder people may get into severe harm to their backbone and that may lead to loss of life. Most of the time in post fall condition the person is lying on ground for many hours once the fall event occurred. This is very significant aspect for person fall detection system to know the seriousness of event. The different techniques are proposed to detect person fall like sensor based, accelerometer and other is camera based. In this paper, two techniques are used to detect person fall such as the deep learning technique convolutional neural networks along with that the image category classification using bad of features is used. The algorithm is providing the promising results as compared to the previously used techniques. The suggested algorithm is described in detail. The algorithm accuracy leads to 100% level on most of the testing measures.