Estimation of Porosity of Microarc Oxide Coating Based on Optical Image Recognition

Abstract

The work is aimed at solving the problem of improving the quality control of coatings with a porous structure. The problem arises due to the lack of an effective and nondestructive method for assessing the porosity of microarc oxide coatings. Accurate porosity control is necessary to ensure the reliability and durability of coatings, as well as to prevent their defects. The use of optical image recognition techniques can improve the process of indirect measurement of coating porosity and improve the quality of control without affecting the object. The factors affecting the porosity of the microarc oxide coating, as well as methods for its determination, are systematized. A method for estimating the porosity of oxide coatings of AD31 aluminum alloy samples based on a recognition program written in the MATLAB R2020a environment is proposed as well as the surface morphology images using modern microscopy methods. A statistical analysis of the surface morphology was carried out, which confirmed good agreement between the porosity estimate and the data obtained during image processing using scanning electron microscope software. The relative error of the proposed method does not exceed 10%. The scientific novelty of the work consists in the development of algorithms for a unique nondestructive testing method, recognition of porous structures based on optical data, which contribute to increasing the efficiency of porosity estimation and improving the characteristics of oxide coatings.