Implementación de un sistema de reconocimiento del uso de mascarillas como medida de precaución contra el covid19 usando deep learning

Abstract

The level of contagion of COVID 19 in the world has reached a critical number for not complying with the correct biosecurity regulations, Ecuador despite being a small country has a large number and to prevent the number of infections from increasing the government together with the The Ministry of Health has indicated various means of biosecurity that the inhabitants must comply with. Using KN95 masks is one of the biosafety methods suggested by the Ministry of Health and in order to control the use of masks, it was proposed to develop a system to detect people who are using and people who are not using masks to thus to be able to keep a stricter control. Artificial intelligence has allowed the human being to be able to create and train neural networks capable of thinking for themselves and making decisions, in the case of this project convolutional neural networks were used that perform the function of detecting which person is Using a mask and who doesn't, with the technology and countless libraries offered by python, opencv and Tensorflow, the neural network was built and trained. To build the neural network, a dataset was built with a quantity of 1500 photographs, which were classified into two groups the first group of photographs of people using masks with mask and the second group of photographs of people who are not using mask without mask.