DESENVOLVIMENTO DE UMA APLICAÇÃO PARA A GERAÇÃO DE UM BANCO DE DADOS DE IMAGENS DE TERMOGRAFIA INFRAVERMELHA

Abstract

Resume. The number of researches in the area of defect detection in composite materials via infrared thermography and the use of deep neurais networks to assist in them has grown rapidly in recent years. With this in mind, and bearing in mind the high cost of thermographic cameras, the present work aims to develop an artificial database with images that resemble the real ones to help and make cheaper the use of these forms of artificial intelligence during academic research on composite materials. For this purpose, the COMSOL software was used, which through simulations, via finite element methods, makes it possible to obtain reliable results if well structured, in addition to facilitating the creation of an application that simplifies the interaction with the model. The project resulted in an application that facilitates obtaining different images to train neurais networks.