Approach of safety-related monitoring and analyzing of human motion with limited sources of captured sensory data

Nowadays, many major efforts and interests are focused on providing better services for some elders and patients who have to live alone, while they are exposed to different dangers during their daily life activities such as falling. The necessity to reduce the negative consequences of falling, whether related to physical or psychological health or to financial encumbrances, by providing an immediate help and a continuous remote monitoring, drives a corresponding meaning to develop a reasonable safety-related monitoring and detecting system of the human motion with respect to the privacy and the simplicity. In this paper, a conceptual prototype of safety-related platform for monitoring human activities by mean of limited number of acceleration and rotation sensors is presented. Additionally, the proposed approach that adopts the artificial neural networks to overcome the challenges emerged by reducing the number of motion sensors and to compensate the missing data is also introduced. Moreover, the required procedures for collecting training and testing data are also summarized in the last section.