Numerous studies have shown that wearing face covers and following other social distancing guidelines reduce community transmission of air borne diseases like COVID-19 during a pandemic. While public health administrators can provide guidelines on wearing masks and social distancing, it cannot be guaranteed that everyone will follow those guidelines. It is very important to have a mechanism available for the public health administration to collect data on how well the population is following the guidelines provided. This information will allow them to anticipate when there will be peak hospitalization and prepare accordingly. This paper aims to look at the feasibility of using machine learning and image processing techniques to track the number of people wearing a face cover in a crowd in real time. The Amazon DeepLens camera, Amazon Rekognition and other Amazon Web Services have been used to make inferences and collect data.