The purpose of this work was to study the effect of roller burnishing process on tribological properties of 36CrNiMo4 steel. The experiments were performed under dry and starved lubrication contact conditions. The influence of selected parameters such as applied load, sliding speed, and sliding distance on the coefficient of friction and wear volume was investigated. The Taguchi technique was used to estimate the parameters significantly affecting the tribological properties. The levels of tested input factors were as follows: applied load—5, 10, and 15 N, sliding speed—0.24, 0.48, and 0.72 m/s, and sliding distance—160, 282, and 404 m. The results revealed that load was the most dominating factor that affects the wear volume and the coefficient of friction. The influence of other input factors was smaller. Regression analysis was also performed to predict tribological behavior. The results demonstrated good agreement between experimental and predicted results. The worn-out samples were analyzed using SEM to reveal the wear mechanism.

The article analyzes the fields of application of machine vision. Special attention is focused on the application of Machine Vision in intelligent technological systems for product quality control. An important aspect is a quick and effective analysis of product quality directly at the stage of the technological process with high accuracy in determining product defects. The appropriateness and perspective of using the mathematical apparatus of artificial neural networks for the development of an intelligent technological system for monitoring the geometric state of products have been demonstrated. The purpose of this study is focused on the identification and classification of reed tuber quality parameters. For this purpose, new methods of identification and classification of quality control of various types of defects using computer vision and machine learning algorithms were proposed.

This article presents a study aimed at using machine learning to automate the analysis of ultrasound images in the diagnosis of polycystic ovary syndrome (PCOS). Today, various laboratory and instrumental methods are used to diagnose PCOS, including the analysis of ultrasound images performed by medical professionals. The peculiarity of such analysis is that it requires high qualification of medical professionals and can be subjective. The aim of this work is to develop a software module based on convolutional neural networks (CNN), which will improve the accuracy and objectivity of diagnosing polycystic disease as one of the clinical manifestations of PCOS. By using CNNs, which have proven to be effective in image processing and classification, it becomes possible to automate the analysis process and reduce the influence of the human factor on the diagnosis result. The article describes a machine learning model based on CNN architecture, which was proposed by the authors for analyzing ultrasound images in order to determine polycystic disease. In addition, the article emphasizes the importance of the interpretability of the CNN model. For this purpose, the Gradient-weighted Class Activation Mapping (Grad-CAM) visualization method was used, which allows to identify the image areas that most affect the model's decision and provides clear explanations for each individual prediction.

This paper presents the theoretical fundamentals, prerequisites for creation, and peculiarities of modeling a new method for determining the refractive index of biological tissues. The method uses a mirror ellipsoid of revolution as an optical element to ensure total internal reflection phenomena. This paper thoroughly analyzes the differences in the refractive index of healthy and pathological tissues on a biometric diagnostic basis. The analysis is used to model the measurement setup’s parameters. This paper also considers various methods of determining the refractive index of biological tissues based on different principles of physical optics, such as interferometry, refractometry, ellipsometry, and goniophotometry. It systematizes typical optical elements of total internal reflection that can be used in goniophotometry. It justifies the selection of the element base for the goniometric installation based on the ellipsoidal reflector method. A simulation of the installation operation was carried out for various parameters of the ellipsoidal reflector, ensuring the measurement of the biological tissue refractive index from 1.33 to 1.7. This paper also proposes a constructive solution for manufacturing an ellipsoidal reflector of the required configuration.

This paper analyzes the light-scattering standards currently used for calibration (verification) and systematic research in photo and spectrophotometry tools. The application specificities in studying the diffuse reflected and transmitted light during biomedical CCD photometry are considered. The advantages of a new class of photometers with non-spherical reflectors as ellipsoids of revolution truncated along the focal planes with the internal mirror surface are presented. The ellipsoid first focal plane is combined with the surface of the under-study media, and the second is optically coupled to the CCD image sensor plane. The principles of zone analysis of spatial distribution reproduced in photometric images on a CCD sensor are substantiated. The illuminance levels of photometric image areas in reflected and transmitted light from the thickness of the standard for the wavelength of laser radiation of 650 nm of different power was experimentally investigated. Polynomial dependences were obtained, and regression coefficients of the illuminance of the external and middle rings in photometric images for the reflected and transmitted light on the laser power were determined.

Timely detection of fires in the natural environment (including fires on agricultural land) is an urgent task, as their uncontrolled development can cause significant damage. Today, the main approaches to fire detection are human visual analysis of real-time video stream from unmanned aerial vehicles or satellite image analysis. The first approach does not allow automating the fire detection process and contains a human factor, and the second approach does not allow detect the fire in real time. The article is devoted to the issue of the relevance of using neural networks to recognize and detect seat of the fire based on the analysis of images obtained in real time from the cameras of small unmanned aerial vehicles. This ensures the automation of fire detection, increases the efficiency of this process, and provides a rapid response to fires occurrence, which reduces their destructive consequences. In this paper, we propose to use the convolutional neural network ResNet-152. In order to test the performance of the trained neural network model, we specifically used a limited test dataset with characteristics that differ significantly from the training and validation dataset. Thus, the trained neural network was placed in deliberately difficult working conditions. At the same time, we achieved a Precision of 84.6%, Accuracy of 91% and Recall of 97.8%.