Desarrollo de un sistema de visión artificial para la detección y secado de humedad en bobinas de papel

Abstract

The objective of the project is to develop a machine vision system to identify moisture in paper reels, which will ensure the quality of the final product. This system uses a thermal camera and machine vision algorithms in images to detect and measure moisture zones in paper reels effectively. Using an innovative method, it will facilitate the detection of moisture patterns that could go unnoticed with conventional techniques. The project design includes the incorporation of four bearings and three guides, distributed in a prototype, to simulate the operation of a paper plant in the post-drying phase. The temperature is measured by the ESP32 with the data from the AMG8333 thermal camera, which is analyzed by a code in the Visual Studio Code application. From this data, the temperature is obtained, which is then transformed into humidity data by means of an execution. The code also indicates when it is necessary to raise or lower the temperature of the nickeline to achieve optimal drying. Paper reels usually break when the humidity is above 60 %, which makes it fragile, but when the humidity is at levels that do not exceed 40 %, the paper is not suitable for commercial sale, for this reason it is important to measure the humidity and have values below 5 %.