Study on signs of defects in the image of the surface of flat-rolled products

Abstract

Currently, more and more challenges of modern industrial enterprises require an increase in the reliability of the information on the quality of products. This becomes possible when using digital technologies to assess the quality of products. The cited publication discusses the technology for recognizing defects in the surface of sheet products in images obtained from cameras of the strip inspection system during rolling. The authors proposed a classification of the signs of defects in the image and highlighted the most significant of them also suggested using geometric, optical and spectral features for images of flat-rolled products containing defects of different classes. The research results at this stage, obtained during the processing of digital images, showed that to identify a defect and reduce false-positive and false-negative alarms of the automated defect identification system, it is required to conduct a study of interval estimates and make decision-making rules based on intersection and merging of intervals; introduce additional classes that allow the introduction of signs that characterize the irregularity of the shape of defects and the characteristic location; the use of new technologies of soft computing will reveal the hidden patterns of the manifestation of defects in the images of the surface of the steel strip.

Grain moisture control in the technological process of drying based on the dielcometric method

“…After the polynomial is calculated, the moisture content of the material is determined and digitally displayed on the indicator (I) and fed to the control object (CO). The technical implementation of the means for measuring the moisture content of grain in the stream requires the development of direct measurement methods implemented by automatic devices (moisture meters) [16].…”