An Image Based Activity Recognition

As the age profile of many societies continues to grow, supporting health, both mental and physical, is of growing vitalness if self-regulating living is to be maintained. Recently, the term of Internet of Things (IOT) has emerged, which highlights the interconnection of all physical and virtually accessible resources with the purpose of collecting and exchanging data. Activity recognition is one of the concepts in ambient assisted living which has been rigorously explored in recent years. The researchers have mainly focused to use sensors to check elderly people. In this paper, we discuss an optical sensor-based monitoring system, for elder people geriatric care and fall detection. In particular, we establish centralized model which works as a decision maker of active recognition. We analyze the photo frame to improve the scalability of our proposed geriatric care system and propose fall detection algorithm to identify the activity of the person.