|
Accepted papers
Neural network-based dynamic grasping of moving objects with robotic manipulators
Yin Cao, A.A. Boguslavsky
Accepted: 2025-11-14
Abstract
We explore the use of the open – source library Stable Baselines3 to implement reinforcement learning via deep neural networks for controlling a manipulator of grasping moving objects along a conveyor belt. Unlike static object grasping, this task requires accounting for dynamic factors, significantly complicating the process. We provide a detailed description of the physical-kinematic modeling of the manipulator in PyBullet and the integration of both the manipulator’s and the moving objects’ parameters into the neural network for training. The results of this study demonstrate that the decision-making capabilities and autonomous behavior provided by reinforcement learning algorithms can be successfully applied to complex tasks, such as dynamic object grasping.
Problems and peculiarities of visualization of traffic flows
, ,
Accepted: 2025-11-13
Abstract
Visualization of transport movement on the plane and in 3-dimensional space is presented. Examples of implementation using frameworks and libraries (WF, WPF, OpenVDB+Blender3D, Manim). A circle consisting of lanes in two-dimensional version and corridors in three-dimensional version was considered as a motion trajectory. The method of setting an arbitrary trajectory for the movement of agents in the model of transportation flow based on the action potential is considered. The formulas allowing to take into account the displacements of both the corridors of the trajectory and the positions of the agents in them are given. This allows us to correctly handle the rearrangements of agents between trajectories, as well as to visualize the simulation results.
Analysis of Data Independence Using Nearest Neighbor Graphs
A.A. Kislitsyn
Accepted: 2025-11-13
Abstract
The article describes a new method for testing the independence of random data sets. This method uses representation of connections between data points in the form of nearest neighbor graphs and compares parameters of the resulting specific graph — such as the number of connected components and vertex degree distribution — to numerically derived critical statistics for random nearest neighbor graphs obtained by the author. The proposed method can be applied to various practical situations. It can be used to test the independence of random vectors in low-dimensional metric spaces. Such problems arise when analyzing data from physical measurements. Additionally, this approach is applicable for analyzing random points in high-dimensional spaces where direct numerical evaluations require exhaustive enumeration and are therefore impractical or impossible to obtain exactly. This problem relates to object classification characterized by many parameters. Moreover, proximity function between points may not necessarily be symmetric, which allows application of graph methods even in such cases. Alongside the task of sample testing, one could also consider comparing pseudo-random number generators by benchmarking structural graph statistics based on them. Considered probabilities of graph structure realization provide an independent set of criteria. For example, the number of fragments in some random graph might be typical for independent random variables while vertex degree distributions could differ significantly. This extends the applicability domain of statistical analysis. The paper presents a collection of model examples illustrating how the methodology works with respect to several types of practical scenarios mentioned above. A comparison of this method with other statistical approaches is provided. We emphasize that using graphs as visualization tools enables immediate identification of dependent elements within samples if there exist cluster centers represented by vertices with anomalously large degrees.
Automated video stream search for research and content analysis of video fragments
N.A. Bondareva, A.E. Bondarev, S.V. Andreev, I.G. Ryzhova
Accepted: 2025-11-09
Abstract
This paper examines the problem of targeted search for specific objects in a video stream on request and recording the timestamps of their appearance. Since the solutions currently available on the market were inadequate for the task, it was decided to implement such a tool independently as part of an ongoing research project conducted at the Keldysh Institute of Applied Mathematics of the Russian Academy of Sciences
Swarm Intelligence: A Quantitative Analysis of Research Publications and Trends
Jeena Joseph, Binu Thomas, L.S. Masalsky, O.S. Logunova, Sabeen Govind
Accepted: 2025-11-05
Abstract
The research investigates the scholarly landscape of swarm intelligence by conducting a comprehensive bibliometric analysis of data sourced from the Scopus database. The search was performed with the help of keywords "swarm intelligence" on the date of 25 January 2024, and eventually, it was found that a total of 1374 articles with diversified sources up to 800 were identified and have been dated between 1996 to 2024. Related dimensions to be studied include annual scientific production, collaboration of publication and bibliographic coupling, leading authors, main sources and affiliations, and keyword co-occurrence. Visualizations used here include line graphs, bubble charts, network diagrams, and three-field plots to present key findings. One can observe from the results that the number of publications increases linearly, although a sharp increase is significantly noticed in 2024. Prolific and influential authors or sources in this area are identified. Moreover, keyword co-occurrence analysis brings out the central concepts or thematic areas cutting across the articles on research in swarm intelligence. A publication collaboration study, such as bibliographic coupling analysis, helps to unveil the international research linked network and the extent to which papers are interlinked. Overall, the understanding that this research has provided me is the tendency, dynamics, and change that has happened in the swarm intelligence research.
On Visualization of Functions in High-Dimensional Space
A.K. Alekseev, A.E. Bondarev
Accepted: 2025-11-03
Abstract
Problems associated with data visualization in multidimensional space are considered. One option discussed is the use of Riemannian space with variable curvature in magnitude and sign for modeling the visualization space. The Hilbert-Einstein, Winslow, and Beltrami equations are considered for modeling the visualization and perception space using geometry. The Beltrami equations can, to some extent, mitigate the problems associated with visualizing multidimensional functions, but are limited by two-dimensionality. The use of Hilbert-Einstein equations is complicated by both the ambiguity of interpreting a priori information and technical difficulties. The most promising approach appears to be the use of Winslow-type equations, which correspond to the construction of harmonic coordinates for the Hilbert-Einstein equations.
The Contours Visualization in Satellite Image of Natural Objects by Artificial Intelligence
E.V. Popov, P.V. Yurchenko
Accepted: 2025-11-03
Abstract
The paper discusses machine learning methods for satellite image classification. We present a neural network algorithm for visualization of the shapes of natural objects, using a variety of machine learning algorithms for preprocessing the training dataset. Our paper compares the classification of the algorithm, calculates its accuracy, and proposes potential improvements. We tested our approach on satellite images of woodland areas in the Bolsheboldinsky District in Russia. The results demonstrate that our improved neural network algorithm achieves high computational accuracy. Robustness, recall, and overall accuracy reach 0.98, especially using training datasets optimized with a support vector machine (SVM). We also demonstrated the applicability of our method for creating accurate geographic information models and detecting changes in natural resources.
Visual and Quantitative Assessment of OpenFOAM Solver Accuracy for Simulating Oblique Shock Train
A.E. Bondarev, A.E. Kuvshinnikov
Accepted: 2025-11-03
Abstract
In the context of the development of computational gas dynamics, selecting the most accurate solver for high-speed flow simulation is a pressing issue. This paper presents a detailed comparative study of four OpenFOAM solvers for simulating the formation of a chain of oblique shock waves. The study focuses on assessing the solvers' ability to accurately reproduce complex flow structures characterized by multiple shock waves. Detailed tables are presented comparing error norms for pressure, density, and velocity magnitude fields. The results indicate that the rhoCentralFoam solver demonstrates superior accuracy. The obtained data can be utilized by engineers and researchers for selecting optimal solvers.
Visual Processing of the Results of Supersonic Flow Around a Delta Wing
T.V. Konstantinovskaya, V.E. Borisov, A.E. Lutsky
Accepted: 2025-11-03
Abstract
The paper considers the problem of visual processing of the numerical simulation results of vortex structures in supersonic flow around delta wing. The methods of visual processing of the obtained structures using various methods of scientific identification of vortex structures and visualization of vortex flows are shown. The obtained results are compared for different incoming flow Mach numbers. The delta wing under consideration had an attack angle of 14°. Numerical simulations were performed on the hybrid supercomputer system K-60 at the Supercomputer Centre of Collective Usage of KIAM RAS.
A Digital Twin Based on Mathematical Modeling of a Wind Turbine Unit
A.E. Bondarev
Accepted: 2025-11-03
Abstract
This article presents a comprehensive approach to creating and visualizing a digital twin of a wind turbine assembly. The study focuses on integrating mathematical modeling with modern scientific visualization techniques. A methodology has been developed that combines multiparameter modeling of dynamic processes with their real-time visualization. Using differential equations of motion and finite element methods, a mathematical model has been developed that takes into account the aerodynamic and mechanical characteristics of the assembly under study. The results of numerical modeling using modern CAE systems are presented, including stress-strain state and aerodynamic analysis. Particular attention is paid to the development of algorithms for visualizing and interpreting multidimensional data.
FakeSegment: Deep Feature Maps Analysis and Visualization for Enhancing of Synthetic Image Realism
V. Aleksandrov, V.V. Kniaz, M. Samodurov
Accepted: 2025-11-03
Abstract
The generation of synthetic datasets for training neural models in object detection and recognition tasks has become a prevalent approach due to the cost and time savings compared to collecting real world data. However, synthetic images often lack critical details, which can degrade the performance of trained models. To address this issue, we propose the FakeSegment neural model, designed to annotate unrealistic parts of synthetic images. Our method utilizes two Single Shot Multibox Detector (SSD) networks with shared weights. By analyzing the differences in corresponding feature maps from real and images. By comparing these feature maps, we can pinpoint areas where synthetic images diverge from expected patterns observed in real-world data. FakeSegment automatically detects unnatural areas within the synthetic data. We evaluate our model on two datasets: the FantasticReality dataset and a newly introduced UnrealLanding dataset focused on aircraft safety during landing. Our contributions include (1) the development of the FakeSegment model, (2) the creation of the UnrealLanding dataset with paired synthetic and real images, and (3) a comprehensive evaluation demonstrating that Fake Segment outperforms baseline methods by 15% in Intersection over Union (IoU) for segmenting unreal parts.
Integration of Generative Neural Networks in Mathematical and Three-Dimensional Modeling: Current State
N.A. Bondareva, A.E. Bondarev, S.V. Andreev, I.G. Ryzhova
Accepted: 2025-11-03
Abstract
This article provides a review of contemporary approaches to the application of generative neural networks in mathematical and three-dimensional modeling tasks. It examines the theoretical foundations of generative neural networks, their architectures, and training methodologies. Existing approaches to mathematical and three-dimensional modeling are analyzed, along with the potential for integration with generative neural networks. Particular emphasis is placed on hybrid approaches that combine the advantages of generative neural networks with traditional methods and expert knowledge, ensuring higher accuracy, reliability, and controllability of results. The article discusses the development prospects and socio-economic implications of implementing hybrid neural network technologies in engineering and scientific domains.
Visualization of Spectral Scenes Using Fourier Series
R. O. Rodionov, E. V. Prikhodko, V. A. Frolov, A. G. Voloboy
Accepted: 2025-11-03
Abstract
This paper presents spectral rendering method that addresses key challenges in storing and processing spectral data. The proposed approach represents light and material properties using truncated Fourier coefficients, allowing spectra to be stored and manipulated compactly. This representation reduces memory usage and computational overhead while preserving the accuracy of spectral information during rendering. The method enables efficient reconstruction of stored spectral functions and simplifies operations such as color conversion. Several strategies for transforming Fourier coefficients within a path tracing framework are investigated, including different spectrum-to-color conversion techniques, such as using zeroth Fourier coefficient to directly convert Fourier-based spectrum to color. Experimental results show that the proposed method provides rendering quality comparable to traditional approaches while producing lower color noise and similar computation times. The method is particularly effective for fast preview and interactive rendering, where low samples per pixel are used and color noise strongly affects visual perception. Also, the paper describes applications of proposed method in neural rendering for storing BRDF using compact neural networks. Furthermore, variance reduction approach based on Fourier coefficients is proposed.
Information Density of Objects in Digital Environment: Theoretical Foundations
N.A. Bondareva, A.E. Bondarev, S.V. Andreev, I.G. Ryzhova
Accepted: 2025-11-03
Abstract
The paper presents a theoretical concept for evaluating information density of objects in digital environment. An analysis of limitations of existing methods for quantitative assessment of information space, based predominantly on simple data volume metrics, has been conducted. The concept of an object's information field is proposed as an aggregate of all informational units containing mentions of the studied object in digital space.
The prospects for applying this methodology in information security, social media analysis, and data preparation for neural network training are examined. The proposed approach opens new opportunities for comprehensive evaluation of information resources and may find application in search engines, recommendation algorithms, and big data analysis systems.
Method of Identification of Implicit Relations in Solving Analytical Problems
I.D. Sokolov, K. V. Ionkina, M. S. Ulizko, E.V. Antonov, E. N. Bazhanova, A. A. Artamonov
Accepted: 2025-11-03
Abstract
Visualization tools enable the transformation of large datasets into user-friendly graphical representations. This paper presents the development of a graph visualization tool designed to uncover implicit relationships among information entities. We describe a method for identifying such implicit relationships and introduce an interactive graph visualization system that allows users to explore the graph through filtering. The implemented functionality includes a specialized query language for dynamically modifying the appearance of nodes and edges.
The proposed method and the developed tool were evaluated on two real-world datasets: (1) detecting potential violations of nuclear non-proliferation commitments and (2) identifying promising areas for scientific collaboration among organizations. The results confirm the practical relevance of the proposed approach.
Development and Application of Machine Vision Algorithms for Workpiece Positioning in Multi-Axis Laser Processing
A.A. Molotkov, O.N. Tretyakova, D.N. Tuzhilin, A.A. Shamordin
Accepted: 2025-11-03
Abstract
This paper addresses the problem of positioning workpieces with curved surfaces for subsequent multi-axis laser processing. The solution is based on recognizing the position of the drawing’s zero point, physically formed by surface height variations of the workpiece. The paper presents an approach to visualizing and detecting the drawing’s zero point using machine vision algorithms applied to a video stream from an industrial digital camera. An algorithm for object boundary detection is described, employing a modified breadth-first search (BFS) with subsequent path reconstruction to the boundary. The developed software module is capable of detecting either the coordinate of a hole center or a workpiece boundary, relative to which multi-axis processing is carried out. In addition, the features of calculating the pixel-to-millimeter conversion coefficient for axis motion are considered, enabling precise movement according to the video channel image. This approach significantly reduces the time required for manual positioning and improves both accuracy and repeatability of the process.
GSDMM Clustering Results Visualization Technique for Short Texts
B.N. Chigarev
Accepted: 2025-10-29
Abstract
The aim of the study is to propose a technique for visualizing the results of short text clustering using the Gibbs Sampling Dirichlet Multinomial Mixture (GSDMM) algorithm, in order to facilitate the analysis of the results and the selection of the hyperparameters of this algorithm and dictionary. GSDMM is selected as the most popular short text clustering algorithm on GITHUB. The algorithm implemented by Ryan Walker on Rust was used. The program Scimago Graphica was used to create bar charts. 16486 bibliometric records on the topic “Visualization” exported from the Scopus database on November 12, 2024 served as the source of short texts. Only Author keywords are used as short texts in this paper. A technique for visualizing the results of short text clustering using the GSDMM algorithm is proposed, which is based on comparing the occurrence of keywords in a given cluster and in each of the other clusters. It is shown that the cluster topics obtained using the GSDMM algorithm can be compared with the results of author keyword clustering performed using the VOSviewer program. The obtained results can be interpreted as a certain stability of cluster themes obtained by essentially different methods. The author suggests to expand the study by creating a thematic dictionary of abbreviations, analyzing the influence of the dictionary on the clustering results of the GSDMM algorithm, and extending the method of visualizing the clustering results to other short texts such as titles and abstracts.
Enhanced Crack Depth Measurement through Binary Image Processing and Geometric Analysis
Kavita Bodke
Accepted: 2025-10-07
Abstract
This paper presents a novel, image-based approach for automated quantifying structural crack width and depth in concrete using binary image processing techniques. Concrete cracks are critical indicators of potential structural failure, and traditional manual inspection methods are often time-consuming, unsafe, and prone to inaccuracies. The proposed method automates crack detection by converting RGB images of concrete surfaces into binary images, isolating the cracks, and measuring their width using the Euclidean distance formula. The depth of the cracks is then estimated using trigonometric relationships based on the measured crack width and viewing angles (30°, 45°, and 60°). This lightweight, cost-effective approach provides a practical alternative to more complex machine learning-based detection methods, making it ideal for real-time infrastructure health monitoring. The results highlight the effectiveness of this technique in accurately measuring crack width and depth across multiple angles, providing critical data for infrastructure health monitoring.
Visual display of tropical cyclone structure zone hazard assessment based on almost periodic analysis
A.A. Paramonov, A.V. Kalach
Accepted: 2025-09-24
Abstract
A visual hazard assessment of tropical cyclone structure zones based on almost periodic analy-sis is proposed. We consider a visualization toolkit that allows us to customize and display both tropical cyclone zones whose radii are multiples of the found characteristic near-periods and shad-ing of the interzonal space taking into account the degree of hazard relative to the cyclone center.
As the main tools we used the visualization modules of the matplotib library Circle and Wedge, which allow us to customize the identified structural zones according to their hazard degree. This development can be useful for emergency-rescue services as an operational diagnostic tool to sup-port decision-making in emergencies caused by natural hazards.
Automated Diabetic Retinopathy Diagnosis and Classification Using Deep Learning with Capsule Network Layers and Stochastic Ensemble Approach
M.A Abini, S Sridevi Sathya Priya
Accepted: 2025-09-10
Abstract
Diabetic retinopathy (DR) remains one of the most common vision-related complications of diabetes and requires timely, accurate diagnosis to prevent severe outcomes. Conventional diagnostic approaches rely on the expertise of ophthalmologists, who manually examine retinal images for lesions—a process that can be time-consuming and prone to fatigue-related errors. To address these limitations, this work proposes a fully automated framework for DR detection and stage classification that leverages recent advances in deep learning. The study focuses on the five recognized stages of DR, ranging from the earliest form, non-proliferative diabetic retinopathy (NPDR), through to the advanced proliferative stage (PDR). The method integrates two powerful pre-trained convolutional neural networks, ResNetV2 and MobileNet, with capsule network layers to enhance feature representation. A stochastic ensemble strategy is applied to further strengthen the robustness of predictions. Experimental evaluation on the Kaggle APTOS 2019 dataset demonstrates a test accuracy of 99.81%, outperforming comparable methods in the literature. Performance was assessed using standard metrics such as precision, recall, F1-score, and the ROC curve. Beyond classification accuracy, the approach also offers improved interpretability through capsule-based visualization techniques and ensemble-driven lesion localization, enabling better identification of retinal abnormalities across different DR stages.
Adapting Virtual Vegetation Samples to Multiobject Visualization of a New Type
P.Yu. Timokhin, M.V. Mikhaylyuk
Accepted: 2025-09-06
Abstract
The paper focuses on the visualization of vegetation arrays within virtual environment systems through a novel type of virtual object known as multiplicable point cloud (MPC). It addresses the challenge of adapting virtual vegetation samples, derived from 3D scans of real natural objects, to the MPC-format. The research highlights key discrepancies between these samples and the MPC-format, and presents a method to resolve these issues using the open-source CloudCompare software. Additionally, the paper proposes an output file format for the adapted virtual samples, utilizing a straightforward and easily interpretable PLY-syntax (Polygon File Format). Under the approbation of the proposed solutions, the adaptation and multiobject visualization of various virtual samples of tree- and grass-like plants were conducted. The results of the approbation affirm the expediency of the proposed method and approach, as well as their potential to effectively enhance the realism of virtual environments.
Principles of organization of continuity in discrete geometrized space
A.V. Tolok, N.B. Tolok
Accepted: 2025-08-24
Abstract
The paper considers the principle of analytical transition to a local function at points on the domain of an implicit function defining a geometric object. Herewith, a transition to partial derivatives is provided to obtain a general form of an implicit local function describing the local geometry for any single point in the object domain. On the analogy of R-functional modeling, a mathematical apparatus for union/intersecting local geometric characteristics of a local function at a single point is provided to construct a discrete region of a complex geometric object.
An example of the intersection of two functions on a defined domain of arguments demonstrates the obtaining of a discretely geometrized three-dimensional manifold for describing a cylinder.
The proposed work is the continued development of method of the Functional Voxel Modeling which offers an analytical structure for the discrete-continuous description of complex geometric objects instead of the means of linear approximation currently used in this method.
Visualization of Complex Roots for Nonlinear Algebraic Equations
Mehmet Pakdemirli
Accepted: 2025-05-26
Abstract
Visualization of complex roots of a nonlinear algebraic equation is discussed in this work. The method is based on calculating the modulus of the complex valued function and representing it as a surface in a three dimensional space where the axis consist of real and imaginary axis and the modulus function. Since it may be inconvenient to visualize multiple roots in a three dimensional surface, contour plot is suggested as an alternative to visualize better the location of roots. Roots of polynomial functions as well as non-polynomial functions are treated as examples. The contour plot is the best to visualize the complex roots in a single graph.
DATA CLASSIFICATION WITH USING VISUALIZATION TOOLS
Andrey Dzengelewski
Accepted: 2025-04-29
Abstract
This article discusses ways to use visualization tools to build object classifiers during automation of a large enterprise. The proposed approaches allow stakeholders to get a visual representation and participate in the decisions required when building a classifier for large arrays of records.
The use of visualization tools is considered when selecting classification objects, determining the attributes and values of classification attributes, ensuring the convenience of the classifier and implementing conflicting requirements from stakeholders. Among the proposed solutions, the methods of using system classes, building logical and physical models of the classifier, multidimensional classification, attribute-value data model, logical data model for describing the required analytics are described.
The subject area is a classifier of works and services, examples of using the proposed solutions and the results of building a classifier at a large enterprise are given.
Using Sperm Imaging with Laser Interference Microscopy for Comprehensive Assessment of the Functional State of Cells during Cryopreservation and under the Action of Molecular Hydrogen
A.V. Deryugina, M.N. Ivaschenko, P.S. Ignatiev, V.B. Metelin
Accepted: 2025-04-29
Abstract
Significant advances have been made in sperm cryopreservation but the search for effective sperm cryopreservation technologies is a pressing issue in modern biology and medicine. The most effective cryopreservation leaves 50-60% of viable cells. The paper discusses the use of molecular hydrogen (H2) as a new approach to enhancing sperm protection during freezing and thawing. H2 is a universal antioxidant and limits damage to biomolecules. Visual registration of spermatozoa under the action of H2 was performed using modern microscopy techniques. Laser interference microscopy was used in the work. Laser interference microscopy records the cell surface architectonics depending on the modulation of the optical density of cellular structures. This visualization option provides information on the metabolic level and expands the possibilities for interpreting experimental results. Sample preparation, dyes, and fixatives are not used in interference visualization. The paper presents an analysis of phase images of spermatozoa during cryopreservation and using H2 as a cryoprotector. Verification of the method for analyzing phase characteristics of spermatozoa as an indicator of the metabolic state of cells was performed by analyzing clinical and laboratory parameters of spermatozoa. The phase height of spermatozoa during cryopreservation decreased, the intensity of energy processes decreased, and the oxidative potential of cells increased. A direct correlation was shown between the phase height of spermatozoa and the concentration of ATP, and an inverse correlation was found from the concentration of malondialdehyde (MDA). The use of H2 determined an increase in the phase height of spermatozoa, an increase in energy metabolism, and a decrease in cell oxidation. Changes in the metabolic activity of spermatozoa under the action of H2 were combined with an improvement in sperm fertility. Thus, phase interference microscopy allows for a qualitative and quantitative assessment of the physiological state of spermatozoa. It is an objective method of vital analysis of complex metabolic activity of cells. It can be used for express diagnostics of their functional state.
Pressure-gradient method for the visualization of a wave attractor
Stepan Elistratov
Accepted: 2025-04-13
Abstract
A wave attractor, a phenomenon of self-focusing of internal/inertial waves on a closed trajectory, has recently been widely studied from different viewpoints. How-ever, there is a lack of investigations concerning its visualization. Peculiar set-ups lately studied show that conventional methods need some improvement.
Herewith, in gas dynamics, the Schlieren method, based on the density gradient, is widely used. Concerning incompressible flows, it is inapplicable; however, pressure can be considered instead density. In this work, a pressure gradient is used as a way to visualize an attractor.
Determination of Adiabatic Wall Temperature in High-Speed Gas Flows Using Infrared Thermography
Í.Ñ. Ìàëàñòîâñêèé, Í.À. Êèñåë¸â, À.Ã. Çäèòîâåö, À.Þ. Âèíîãðàäîâ
Accepted: 2025-03-21
Abstract
This paper presents a method for the non-contact determination of the adiabatic wall temperature in high-speed gas flows. The method is based on the processing of a sequence of thermograms obtained using an IR camera, within a program developed in Python 3.10. The approach demonstrated high efficiency when handling large datasets, particularly concerning minimizing temporal and computational demands. The adiabatic wall temperature was determined under both steady-state conditions, directly in the experiment, and transient conditions, through the extrapolation of the heat flux as a function of the current temperature of the examined surface. The effectiveness of this method was demonstrated in the investigation of non-mechanical energy separation in compressible gas flows.
Three-dimensional images of residual strain fields by wavelet transform method
I.V. Laktionov, E.V. Gladkih, A.P. Fedotkin, G.Kh. Sultanova, À.S. Useinov
Accepted: 2025-03-21
Abstract
The accuracy of measuring Vickers hardness values depends on image focusing both during automated determination of residual imprint diagonal lengths and during operator working. Widespread algorithms for image focusing are based on brightness and contrast adjustment. We propose a new approach based on alternative algorithms for more accurate microscope focusing system used in marking imprints after indentation. Implemented algorithms are based on variance, Laplace function and wavelet transform. We select the optimum values of the basis and transform depth when using the wavelet transform. We tested new approach on samples with poor contrast, rough surfaces, and materials with pile-ups occurred in the indentation process. Applying different focusing functions depending on focus position demonstrates a more stable performance of the algorithm with wavelet transform. We also demonstrated obtaining a fully focused frame and a pseudo three-dimensional map of the sample.
Use of Hadamard matrices in single-pixel imaging
Denis V. Sych
Accepted: 2024-08-13
Abstract
Single-pixel imaging is a method of computational imaging that allows to obtain images of objects using a photodetector that does not have spatial resolution. In this method, the object is illuminated by light having a special spatio-temporal structure, — light patterns, and a single-pixel photodetector measures the total amount of light reflected from the object. The possibility of obtaining an image and the image quality are closely related to the properties of the applied patterns and computational algorithms. In this paper, we consider patterns obtained from modified Hadamard matrices and study the features of image calculation using single-pixel imaging. We show the possibility of reducing both the sampling time and the computational resources required to obtain images by modifying the pattern system. The proposed theoretical method can be used in the practical implementation of the single-pixel imaging method in an experiment.
|
|
|
