In the process of manufacturing metal products and metal structures, various defects often appear in the material that affect the performance characteristics of the object. In order to detect internal defects, a radiographic control method is used, which allows assessing the quality of the material of the test object from an X-ray image, which is further digitized and transmitted for analysis. The article discusses the main methods of searching objects on a halftone image, and also describes the process of developing an information system using the Haar cascade classifier as the main image processing tool.