The image obtained in metallography is an image of the structure of the material on which grains of different phases are present, nonmetallic inclusions, etc. Quantitative parameters are determined for the elements of the structure by the methods of quantitative metallography. In the past, the quantitative evaluation of the structure was carried out on the basis of the reference structures contained in GOSTs, as well as by primitive measurements. Modern image processing programs allow you to determine the area and the linear grain sizes of any phase, perimeter, and also the derivatives of these values- the shape factor, the average size, the equivalent diameter, etc., and also perform mathematical processing of results, build a variety of mathematical relationships - the distribution of a certain value (grain size, diameter, etc.) in size (frequency curve). In this case, it is possible to generate files in Word and Excel containing tables with the received data, as well as graphs constructed from this data. Determining the area of objects in the image processing program is the most objective type of analysis. The area definition in the image processing program is the summation of the areas of all the pixels that make up the object (belonging to the object). Image processing programs also determine, for example, the maximum and minimum diameters, as well as the length and width of the object. There is some conventionality in the definition of these parameters. The minimum and maximum diameters are the lengths of the sides of the minimal bordering object of the rectangle, whose sides are parallel to the boundaries of the image. Length and width are the lengths of the sides of the bound rectangle bordering the object, whose sides are parallel to the axes of symmetry of this object. A specific set of defined parameters can differ in different image processing programs. The question of the measurement error in estimating the quantitative parameters of the structure is raised from time to time in discussing the results of scientific work and is important for explaining the results of the research. As a rule, the question of an error is not discussed if the metric segment is the accessory of the microscope objective and is fixed together with the image of the structure. Some microscopes do not have such a metric. The image processing program for this case should include a calibration table of magnifications. To form it, a standard object-micrometer is used to work in reflected light. Errors in measuring the size of structural components are made at the stage of image calibration, as well as when objects are detected. The main calibration error is determined by the apparent thickness of the object-micrometer stroke, which changes with increasing magnification. Also, the error is related to the decrease in the number of divisions with increasing magnification, for which data are generated for a calibration table of magnifications. The error in determining the grain area at this stage varies from 0.05% with an increase of 100 times to 1% with an increase of 2000 times. In metallography, the error is also associated with the subjective error of the operator (the character of the hand movement, visual acuity, various qualifications, etc.). When multiple processing of the same frame by one operator, the relative error in determining the grain area is ~ 0.9%. When processing the same frame by four different operators, the greatest relative error associated with the unequal perception of the image of the structure by different operators is about 6.0%.