Deep Learning Based Face Mask Detection System in OpenCV

Now a days, a life threatening disease called COVID -19 is spreading very rapidly from 2019 onwards. It also affects the respiratory system of human very badly that leads to a horrible death. This death ratio is increasing all over the world, day by day. In India, death rate has reached upto 4.03 lakhs. The World Health Organization (WHO) is suggesting the common people to follow the general instructions like wearing the mask properly at public places; sanitize the hands at regular intervals and maintain social distancing at public places. Along with this, Indian Government and other countries are initiating the immunization process and giving the awareness about the immunization. But still, people should wear the mask in common places to cut over the chain of corona virus and its new versions. In keeping this view, very few researches has done till now and there is a need to explore more researches to detect whether the people are covered their face with the mask properly or not in the common places. There are two phases in proposed system, out of which, first phase is to train the system and second phase is to apply the system to detect the faces. Keras / Tensorflow, OpenCV and MobileNetV2 have used to achieve better results and this model has been trained with the help of large dataset that includes the images of faces with and without mask. By using this architecture, the manufacturing cost of this face mask detection system is reduced to some huge values.