With the advancement in technology, personalised monitoring is emerging. During the COVID-19 pandemic, the old age homes have the most susceptible population. We have used activity recognition to detect abnormal activities which also include symptoms of COVID-19 like sneezing, fatigue and headache. Elderly people are prone to falls which can be fatal or cause injuries. Activity recognition and fall detection can be used to keep a tab on them. Generally senior citizens live alone or in care homes. Hence if they are constantly monitored, adequate help can be provided. In this paper, we propose a system with activity recognition and fall detection. OpenCV is used for activity recognition and Markov model for the sequence of activities. The probability of that sequence of activities is predicted. If it is below a threshold, it is considered as an abnormal activity. A Single Shot Detector model is implemented for fall detection.