Automated Method of Monitoring the Quality Surface of Glass And Mirrors by Means of Machine Vision Algorithms for Gyroscopic Devices

Abstract

The paper analyzes the existing devices for surface quality control of products. A variant of machine vision system application in terms of products using glass surfaces with high requirements to the quality of manufacturing are considered. Ready-made solutions are considered and analyzed, their characteristics, advantages and disadvantages are described. The layout of a workstation for obtaining images of the product surface for subsequent software processing was developed. An example of automatic acquisition of surface images for products with complex geometrry is implemented. An algorithm and tool for detecting surface defects and classifying them, by image processing, using specially developed software, is described. The described method of surface assessment is also capable of detecting defects in various surfaces requiring high quality production of these products, and the main limitation is the optical method of obtaining a picture of the surface to be examined. The algorithm is able to work with images of medium and high quality, which allows to use this method of surface assessment with already existing systems that do not have an algorithm for surface assessment, which simplifies the possibility of implementing this system. Combination with other methods of nondestructive testing of products, that allow to obtain surface images, internal area and hidden cavities, it will make possible to analyze the quality of products in a complex way. The ue of machine vision is relevant due to its features, such as: speed of surface quality control, as well as evaluation accuracy. A machine vision system can reject parts much faster and more reliably than a human and can be used as a tool production automation.