Falls are a fatal threat to the elderly peoples health. It is a notable cause of morbidity and mortality in elders. Falls can even lead to serious injuries and death of the person , if they are not given proper attention. Above 30% of persons aged 65 years or above , fall each year and they mostly are reoccurring. The severity of such falls are due to the increasing age, cognitive impairment and sensory deficits. A multidisciplinary approach should be developed to prevent future falls. This paper emphasizes the need and development of an advanced fall detection system using Machine Learning and Artificial Intelligence technologies. The fall detection systems are currently categorized into wearable and non-wearable devices existing in the market. These wearable devices use sensors which may not be accurate always and it would be difficult for the elderly person to wear it around their body all the time. The architecture that is proposed in this paper uses open source libraries such as OpenPose for a much better detection and alert system, among non-wearable devices. The system retrieves the locations of 18 joint points of the human body and detects human movement through detecting its location changes. The system is able to effectively identify the various joints of the human body as well as eliminating environmental noise for an improved accuracy. This results in improved effective training time as well as eliminating blurriness, light, and shadows. The developed approach falls within the scope of computer vision-based human activity recognition and has attracted a lot of interest.

Keywords : OpenPose ; OpenCV ; Fall detection ; Artificial Intelligence ; Human Action Recognition ; Convolutional Neural Networks ; LSTM ; Image preprocessing ; Recurrent Neural Network