Computer Vision Based Wellness Analysis of Geriatrics
M. Sivakumar, E. Iswarya, K. Malusha, T Priyadharshini · 2021 5th International Conference on Electronics, Communication and Aerospace Technology (ICECA) · 2021
The evolution of modern technologies has made real-time and personalized monitoring immensely helpful to the geriatrics. With the increasing trendin the growth of the elderly population, it has become cumbersome for caregivers to keep a check on seniors' day by day exercises physically and ceaselessly. Trends in their daily activities, such as eating, sleeping, sitting, standing, walking, drinking etc. can provide caregivers proper information with regard to seniors' health. Hence, the proposed idea is to devise a self-operating model to continuously scan the seniors' activities and to provide them with descriptive analysis using Image Processing techniques and Deep Learning algorithms. The proposed methodology consists of the Human activity recognition and the classification of the Human activity using ResNet-101, VGG-16 and Inception V3 which are Convolutional Neural Network Architectures and to propose the best model based on its performance. The model will detect the elderly's activity, recognize the activity and finally provide the descriptive wellness analysis.