Monitoramento de produção de torradas em uma esteira industrial por visão computacional

Abstract

Industry 4.0 symbolizes the union of the industry with the technological advances that are currently happening, such as artificial intelligence, internet of things and robotics. Computer vision is one of the ways that the industry can modernize itself, using cameras to monitor its processes and make decisions based on the data obtained from the analysis of the images of the processes. In this work, it is presented a computer vision system that uses digital image processing techniques for monitoring the production of toasts on a belt conveyor, making it possible to automate responses to events that may occur in a production line. For the system validation, videos were made of simulations of toasts being transported in a belt conveyor, this way there is a controlled test environment, where the angle of rotation and size of the toasts are known. Satisfactory results were obtained, where the system was able to detect all the toasts, and the measured rotation degree of the toasts had a maximum error of three degrees, showing that this system was effective in detecting and analyzing the toasts presented at the production line.