Authors return to the topic of the article presented at Graphicon-2019 and review the issues of recognizing material microstructure elements, most common problems, recognition errors, and resolving the issues. Authors present methods for reconstructing particle shapes after the removal of image artifacts and separation of touching particles. The article includes examples of applying developed algorithms for analyzing particles present in graphite microstructure in cast iron and for analysis of powder particles. The article also demonstrates the possibilities of applying quantitative analysis methods for images, including defects.

Analysis of metals' microstructure images is an actual quantitative analysis problem, solved by quality control and research labs in the field of metallurgy. SIAMS Ltd pursues the goal of improving microstructure analysis quality, speed, and convenience. This article discusses the issues of recognition of the microstructure elements of metals and alloys, the most common problems and recognition errors, methods for solving them. Microstructure examples are given before and after digital image processing. The question of the advisability of using server technology in digital microscopy that removes such restrictions from users as the size of the shooting area and the area of microstructure analysis, binding to one working computer and one software license, as well as restrictions on the exchange of results between industry experts, will be raised.