Please use this identifier to cite or link to this item: http://hdl.handle.net/10609/132653
Title: Control de accesos con detección de mascarillas en tiempo real implementado con algoritmos de reconocimiento facial y Deep Learning
Author: Cuadra, Ezequiel
Tutor: Sansano Sansano, Emilio
Others: Arnedo-Moreno, Joan  
Abstract: Due to the pandemic we are suffering because of covid-19, governments have had to adopt a series of restrictions and preventive measures to avoid the spread of this virus. The purpose of this project is to create an access control system with real-time mask detection, implemented with Deep Learning algorithms designed for computer vision. Through Transfer Learning techniques we will take advantage of a pre-trained network with a great computational power and on a huge image dataset (Imagenet). We will freeze the first layers and create a network that will be coupled to the head of this model to train with our data. This system can be used in different environments to ensure that security measures are being properly enforced, such as factories, public transportation, enterprises and more. It is concluded that a network previously trained in image recognition is able to recognize faces and classify properly if a mask is worn, if it is worn correctly or if it is not worn at all. This is achieved by training and coupling our network with the previously trained model. The development of the project will be carried out in Python language, with support libraries such as Keras for neural networks, Pandas for data manipulation, Numpy for mathematical functions, OpenCV for computer vision, and others.
Keywords: deep learning
control access
masks detection
facial recognition
Document type: info:eu-repo/semantics/bachelorThesis
Issue Date: 8-Jun-2021
Publication license: http://creativecommons.org/licenses/by-nc-nd/3.0/es/  
Appears in Collections:Bachelor thesis, research projects, etc.

Files in This Item:
File Description SizeFormat 

mask_detector_1dataset.mp4

28,46 MBMP4View/Open
ezequielmejorTFG0621memoria.pdfMemoria del TFG1,52 MBAdobe PDFThumbnail
View/Open