Covid-19 face mask detection using convolutional neural network and image processing

Mengistie, T. T.; Kumar, D.

Object detection and image classifications are the popular research work in the rapid growth of advanced technology to detect and identify real-time problems in the governmental core-areas such as business shops, airlines, universities, and army place with the help of webcams surveillance cameras, and opensource platforms. The objective of this research work is to propose Convolutional Neural Network (CNN) and Open Source Computer Vision (OpenCV) to detect COVID-19 face mask detection from real-time (live streaming video) and image datasets. We have employed data argumentation pre-processing techniques to maximize our dataset and improved the performance of the proposed models, and in addition to this, we have used MobileNet CNN architecture to reducing the size and model complexity. This research work's experimental result has been shown 96.50%, 97.07%, 99.27% accuracies obtained with the help of CNN, CNNdata augmentation, MobileNet2V, respectively. © 2021 IEEE.