Implementation of Real-time Face Mask Detection with Convolutional Neural Network (CNN) and OpenCV

Abstract

AbstractIn this global pandemic of COVID-19, there is a critical need for self-protective devices, and the most important of them is a face mask. Our project’s main aim is for identifying the presence of a face mask on person's face. A strategy should be formulated to make the people accept this essential safety measure. To check this, a face mask detector system should be used. To check the presence of a face mask on a human face, the primary step is the detection of human face. This can be divided into two parts: verifying faces on the images and detection of masks on their faces. Face masks are used to prevent cross-contamination as part of an infection control strategy. Using TensorFlow, Kera's library, and OpenCV, we created a very rudimentary convolutional neural network (CNN) model. Our experiment demonstrates that it operates effectively on test data, having a precision of 100% and a recall of 99%, respectively.KeywordsCOVID-19Cross-contaminationConvolutional neural network (CNN)Face masksComputer vision