Real-Time Fall Detection by combining features based on machine learning

The world's elderly population is growing every year. Falls are one of the biggest dangers for older people living alone at home. This paper presents a fall detection Model to support the independent living of the elderly in an indoor environment. The aim of this paper was to investigate appropriate methods for diagnosing falls through analyzing the movement and shape of the human body. Serval machine learning Technics have been proposed for automatic fall detection. Existing fall detection technologies fall into three main categories: computer-based techniques, ambient device-based techniques and wearable sensors. The proposed research reported in this paper detects a moving object by using a background subtraction algorithm with a single camera. The next step is to extract the features that describe the change in human shape and recognize the differentiation of falls from activities of daily living. These features are based on motion, changes in human shape, and Oval diameters around the human and temporal head position. The features extracted from the human mask are eventually fed in to various machine learning classifiers for fall detection. Different machine learning methods were compared to evaluate their ability to accurately detect falls. Experimental results shown the efficiency and reliability of the proposed method with a fall detection rate of ۸۱% that have been tested with UR Fall Detection dataset.