Face Mask Detection to Ensure Covid-19 Safety by Using OpenCV, Tensor Flow and Keras

Abstract

We are living in a pandemic period. Covid-19 has changed our daily life in great manner. The economy of the world affected largely for the pandemic. We are under great health crisis. Thousands of people are being affected by this in daily basis. Droplets of an affected person can spread the virus among the people around him or her. Thus, an affected person is threat for others. As there are lots of people in public places, so the chances of transmission of the virus is very high in public places. From the world health organization we have known that Covid-19 can stay 14 days on a host without showing any symptoms. In this long time, the host may spread the virus among a community without his conscious. Again, another 48 hours is required to confirm if a suspect has covid-19 or not. But in this circumstances, 48 hours is also a long time. Many people can be affected in this time and then they will also spread the virus among millions. According to WHO, at this moment wearing a safe facemask is the best way that can stop the spreading of Covid-19. If people wear mask, their droplets will not spread in the air and people around him will be safe. We see various service providers refuse to give any services if a customer do not put mask in his face. So, detecting safe mask on person’s face is become a highly appreciated work. In this project, we also going to do build a system that can perform this great task. Here we will use some basic deep learning and machine learning technique to perform the task. We will also need help of some library packages of python such as OpenCV, Keras and TensorFlow. Using all of this we hope to build a system that will detect if a person is putting facemask on his face or not correctly.