A Novel Approach for Facemask Detection and Alert System Using Machine Learning

The Covid-19 pandemic has changed the whole world. It has effected economically and socially and has changed the human lives unexpectedly. It spreads from the infected persons and is contagious. We can reduce the exposure to the virus using facemask and social distancing. Although government has given outrules that wearing facemasks and keeping social distance is obligatory because of the social inattention, numerous are still being affected by the virus especially the transmission of virus in closed rooms or indoors in more when compared to outdoors. Wearing of facemasks incorrectly is one of the major problems nowadays, and monitoring each and every person all the while is a bit tidy task. As the technology is advancing we have latest advancements in the field of artificial intelligence and face recognition techniques that can be used to identify the persons who are wearing mask incorrectly and reduce the virus transmission. In this paper, we recommend an approach to automatically identify the persons who are wearing the masks incorrectly. We us OpenCV and Keras to identify the persons wearing facemask incorrectly and notify the authorities about the person not wearing the facemask and stop their entry at the entrance itself. We can implement this approach in the surveillance cameras, and thus, this can be implemented without requirement of any additional hardware. We have achieved high training accuracy of 96.26% and validation accuracy of 92.59%, respectively, on two different datasets than our earlier models and implemented this model successfully with alert system.