Software for research automation in the field of materials science

Modern approaches to automating the analysis of the microstructure of metallic materials aimed at improving the accuracy and efficiency of research are presented. The development of software (SW) for the identification and classification of grains in metals is described, which is a key aspect in studying their structure and predicting mechanical properties. The program includes modules for partially automated image processing, grain characteristics analysis, visualization of results, and integration with machine learning algorithms. Specialized tools allow you to identify grain boundaries, analyze their size, shape, orientation, and automatically calculate pixel sizes for accurate analysis. The software is developed in the Python programming language using the OpenCV, NumPy, and Scikit-Image libraries, which provides extensive opportunities for further implementation of adaptive machine learning algorithms. The main stages of the program include downloading and preparing images, highlighting the boundaries and contours of structural elements, image segmentation, and analyzing grain characteristics. The results of the analysis are presented as an automatically generated report. The implementation of an automatic scale line recognition system in microphotographs is described, which makes it possible to accurately determine pixel sizes and, accordingly, grain sizes. This is important for improving the accuracy of calculations and analysis of metal microstructure. The developed SW is focused on applications in scientific research and industry, such as quality control of metal materials, optimization of thermomechanical processing processes and creation of materials with unique properties.