Features of metal structures digital images containing carbides investigation

The analysis of microsections requires the involvement of highly qualified experts in the field of materials science, which, in turn, does not exclude the influence of the "human factor". On the other hand, the issues of increasing the objectivity of identifying the properties of metals and alloys require the use of modern data processing methods, for example, artificial intelligence in solving problems of classification and identification of macro and micro structures.The paper presents an overview of studying macro and micro structures containing carbides process, determining the specific features inherent in these images, and proposing an information model for their processing. The article is devoted to the development of an information model intended for the analysis of metal structures digital images with carbide inclusions. The analysis of literary sources is carried out, it is established that the study of metal structures is an important tool for assessing qualitative characteristics. The presence of carbides in the metal structure has a significant impact on its quality. A review of the methodology for studying the structure of a metal is given, and the importance of metal structures image processing stage is determined. The main methods for obtaining digital images of the alloy structure are described. Samples of metal structures with carbides are presented. A procedure for digital processing of metal structures images with kibide inclusions is proposed, which consists of image conversion to grayscale, contrasting, and threshold binarization. An analysis of the results of metal structures processing images made it possible to identify areas with carbide inclusions, however, additional artifacts that were not carbides were found in some images. Balancing by the binarization threshold in this case does not improve the detection of carbide inclusions network due to the lack of contrast. Histograms demonstrate the presence of information features in a wide range of gray colors, so for this class of images, more sophisticated image processing technologies need to be developed. In the course of digital images features study of metals and alloys metal structures containing carbides, it was: an information model for processing metal structures containing carbide inclusions is proposed; the proposed information model is applied to digital images of metal structures; it was found that some images of metal structures are characterized by low contrast, which leads to the selection of background artifacts, except for areas with carbide inclusions; the development of complex mathematical methods for the detection of carbide inclusions in images of metal structures characterized by low contrast is proposed. Thus, the article shows the results of carbide inclusions of the using the digital image processing procedure. The advantages and disadvantages of the approach are shown, the directions for its improvement are determined.