Tempering color classification via artificial neural network (ANN): An intelligent system approach to steel thermography

• Main Authors: Cotoco, Ezekiel Karl A., Lindo, Delfin Enrique G., Baldovino, Renann G., Dadios, Elmer P.
• Format: text
• Published: Animo Repository 2018
• Subjects: Steel—Heating; Neural networks (Computer science); Manufacturing
• Online Access: https://animorepository.dlsu.edu.ph/faculty_research/1921
• Institution: De La Salle University

In our modern society, the steel industry is a critical component to achieve economic growth and development especially in the infrastructure and manufacturing industries. However, steel production is not just an easy step process. Untempered steel, though hard, is too brittle to be useful for most applications. In order to enhance its properties, the application of heat treatment is performed to steel. Heat treatment is a meticulously sensitive and an extremely tedious process due to temperature sensing. Nowadays, the common way to determine the temperature of a certain metal is through the use of human vision or a thermal imaging camera. However, these methods are either inaccurate or very expensive to setup. In this study, the application of artificial neural networks (ANN) in assessing the steel discoloration when it undergoes extreme temperatures is a cheaper and more accurate way of reading or sensing its temperature. The use of neural network technology can easily adapt to classify a wide range of discoloration from different metals especially steel. © 2017 IEEE.