Detecção e determinação da porcentagem de pessoas usando máscaras faciais em imagens capturadas por câmeras fixas ou drones para auxílio na prevenção do COVID19

• Main Author: Freitas, Caio Henrique
• Publication Date: 2021
• Format: Bachelor thesis
• Language: por
• Source: Repositório Institucional da UNESP
• Download full: http://hdl.handle.net/11449/234899
• topic: Machine learning; Deep learning; Convolutional Neural Networks; Image recognition; Aprendizagem de máquina; Aprendizagem profunda de máquina; Redes Neurais Convolucionais; Reconhecimento de imagem; Redes neurais (Computação); Processamento de imagem assistida por computador; Transmissão de imagem

The spread of Corona Virus around the world, took away millions of lives. The use of facial mask is one of the best ways to stop the spreading process of the virus, as long as social distancing, and they are widely disclosed by World Health Organization. This research work has the objective of developing a software capable of detecting the use of facial masks, by analysing images from surveillance cameras and drones. It uses VGG-16, YOLO and ISR algorithms, and it is capable of defining a percentage of people using mask. The architecture used has reached about 87% accuracy for the used validation dataset