Automatic fall detection for elderly-providing timely aid using GSM modem

The number of persons above the age of 60 years is fast growing, especially in India. Falls are one of the major problems in the elderly and are considered one of the "Geriatric Giants". In the paper, we propose a camera based automatic fall detection method to facilitate the independence of older adults, reduce risks, and enhance the quality of life. The proposed approach consists of three parts: object detection, fall confirmation and reporting the problem. An adaptive background subtraction method is used to detect a moving object. A set of features such as aspect ratio and fall angle are extracted from the obtained foreground image to analyze, detect and confirm a fall. Our approach is unique in a way as each person in the video is separately monitored (when there are multiple persons) for fall detection. As soon as the fall is confirmed, it will be reported to the nearby hospital service with the help of GSM modem (TTL logic). To facilitate serial communication, USB to TTL logic convertor has been used eliminating the need to use microcontroller, making it simpler and cost effective. Key words: Background subtraction, fall detection, GSM Language: en