When designing lens systems for optical sensors, a fundamental operation is finding the intersection of a ray and the lens surface. We propose a robust method for finding the intersection of a ray with highly aspherical rotationally symmetric surfaces. Our method provides reliable results for ray incident at large angles relative to the optical axis and successfully calculates intersection points for a significantly larger number of rays than existing methods, improving the reliability of optical modeling and design. The core idea of our method is to calculate an initial approximation of the intersection point for the Newton method using a ray marching procedure with a step size estimated form of analytical derivatives on the surface function. The proposed method does not require maintaining additional complex data structures and is suitable for use in the iterative optimization of optical systems where the surface geometry frequently changes.

Natural lighting (daylight) plays a decisive role in energy saving and building maintenance. An important aspect is the comfort of the premises. Modern building design requires computer analysis of indoor illumination in accordance with the green building certification program, the main indicators of which are spatial daylight autonomy (sDA) and annual sunlight exposure (ASE). Calculation of these metrics requires thousands of simulations of global illumination for various time moments throughout the year. Well-known methods of lighting simulation are not effective here. We propose an original approach that uses precomputed data and its fast interpolation on a Klems grid to iteratively simulate the lighting distribution. The method was compared with classical stochastic ray tracing and existing solutions based on the radiosity method. The efficiency of calculations has increased significantly (and sometimes orders of magnitude) while maintaining high accuracy of the results. The ability to calculate metrics in tens of minutes on a regular computer, provided by our method, allows an architectural design to be continuously checked for compliance with the green buildings program.

Viewpoint selection methods for 3D scenes have various applications in computer vision, scientific visualization tasks, and in 3D reconstruction to reduce the training sample size. In this paper, we propose a method for viewpoint selection in application to material reconstruction problems based on inverse rendering using modified viewpoint quality scores as a loss function. The proposed method obtains viewpoints based only on the geometric characteristics of the 3D object and constraints on the camera positions and does not require some pre-known set of viewpoints. The proposed method is tested on an open dataset for 3D reconstruction.

Instrument panels of modern aircrafts are created based on the concept of glass cockpit. This new interface philosophy improves the perception of important flight information by displaying it on a single multi-function display. This paper considers problems that arise when developing a certified cockpit display visualization system designed for operation on civil aircrafts under the Russian real-time operating system JetOS. The paper presents several algorithmic solutions that make it possible to achieve acceptable rendering speed. In particular, a solution to the problem of rigid scheduling of operating system partitions is described in detail. This solution makes it possible to overcome the degradation of rendering speed. Some directions for further research are outlined.

Currently, virtual prototyping, i.e. preliminary modelling of processes on a computer has become ubiquitous. It is also applied to modern automobile paints in order to correctly reproduce the colour of the paint. The most practically important task is colour matching: calculation of the paint composition necessary to obtain a given colour. The complexity of modelling the propagation of light in paint is due to its complex structure and the wide variety of pigments, including diffuse ones. This problem can be solved by simulation of the light scattering by the dispersed pigment particles. To do this we need the optical properties of pigments, which are often unknown. In this paper, we propose several different methods for extraction of the pigment properties by measuring the light scattering of specially prepared samples. All methods are based on minimizing the difference between measured and calculated scattering values as a function of pigment properties. The methods differ in the types of paint samples that must be prepared and measured to obtain the pigment data. We describe the operation of the methods and test their robustness. As a result, one method has proven to be much more robust to the measurement errors. The colour of the real sample was also compared with the calculated colour of the paint, in the modelling of which the obtained properties of the pigments were used.

The main computationally expensive task of realistic computer graphics is the calculation of global illumination. Currently, most of the lighting simulation methods are based on various types of Monte Carlo ray tracing. One of them, the Langevin Monte Carlo ray tracing, generates samples using the time series of a system of the Langevin dynamics. The method seems to be very promising for calculating the global illumination. However, it remains poorly studied, while its analysis could significantly speed up the calculations without losing the quality of the result. In our work, we analyzed the most computationally expensive operations of this method and also conducted the computational experiments demonstrating the contribution of a particular operation to the convergence speed. One of our main conclusions is that the computationally expensive drift term can be dropped because it does not improve convergence. Another important conclution is that the preconditioning matrix makes the greatest contribution to the improvement of convergence. At the same time, calculation of this matrix is not so expensive, because it does not require calculating the gradient of the potential. The results of our study allow to significantly speed up the method.

The work proposes an approach to programming parallel architectures without using parallel programming interfaces (APIs) and parallel directives for various numerical simulation and computer graphics applications. The primary goal of this approach is to resolve the fundamental conflict between cross-platform compatibility and hardware acceleration when developing high-performance programs. This conflict is resolved through the automation of the development process: algorithmic descriptions in C++, free from any specific parallel constructs, are automatically translated into the implementation of the same algorithm, and a realization is generated on an existing parallel architecture programming API (C++ translates to SPIR-V for GPUs, and vector instructions for CPUs). Furthermore, if specific hardware acceleration is required, the programmer can replace individual parts of the generated code by substituting kernels and virtual functions in the generated code. Such substitution allows for code regeneration without losing usermade changes tailored to a particular hardware platform. The developed system operates as a white box, allowing the programmer to read and debug the generated code as if it were written manually. This enables us to easily distinguish translator errors from user errors and, moreover, does not create dependencies on the developed system in projects. The generated code can always be manually rewritten when necessary. The application of the developed technology is considered using the example of one of the most complex and inconvenient problems for GPUs — parallel algorithms for constructing trees.

The work is devoted to the application of previously developed algorithms and methods of stereo animations construction in the field of biological research on the example of stereo images for functional tomogram of the brain. The construction of stereo images on the autostereoscopic monitor provides an opportunity to obtain an in-depth understanding of the studied object. The autostereoscopic monitor allows viewing stereo images without glasses, while providing a quality not inferior to that of a classical 3D projection stereo system. This work continues the cycle of research conducted at the Keldysh Institute of Applied Mathematics. The specificity of the displayed data allowed to improve authoring libraries for creating stereo images. The research is aimed at developing technologies for constructing stereo images and animations for presenting the results of scientific calculations on classical stereo devices and autostereoscopic monitors. Constructing a stereoscopic visual representation of the results of biological research will enable researchers in the field to gain a deeper understanding of the object under study and its properties.

Images in vector format are presented as a sequence of analytical descriptions of geometric objects. This approach allows for reproduction of the image in any resolution without loss of quality. Currently, there are no ready-made solutions for using vector images in GPU photorealistic rendering systems. This paper presents an approach to enabling such support using signed distance fields and rasterization as base methods. Analysis of the results shows the effectiveness of the approach based on distance fields for various vector images. However, in some cases, artifacts may appear, in which case it is proposed to use a rasterization-based approach.

Nowadays, computer simulation is being used to develop new materials. Many of them are dispersed media (e.g., paints, and 3D printer inks). Modern automotive paints are of great interest in research works. They contain colorant particles and thin flat metallic or pearlescent flakes distributed in a clear varnish. There are two main approaches to simulation of light scattering in a dispersed media. The first one is based on the continuous medium model. This model is faster but less accurate. The second approach is the simulation of light propagation through an ensemble of paint flakes and particles represented as an explicit geometry. This model correctly calculates light scattering but is rather time-consuming. In our study, we investigated the dependence of the painted surface luminance on particle size and compared both the approaches. We prove that the effect of coarse particles can emerge even in a model where positions of these particles are not correlated; this is different from the mainstream studies which have only concentrated on the role of these correlations. Then, we suggest a semi-analytical model of dependence on particle size. This model not only allows to more accurately simulate visual appearance but also admits intuitive comprehension of how it is affected by various medium parameters. In case of the divergence between the results of LTE and accurate approaches, we propose a simple approximation that allows to improve the accuracy of the LTE results for coarse particles.

The instrument panels of modern aircraft are created according to the concept of a "glass cockpit". This new interface concept enhances the perception of critical flight information by integrating it into one multi-functional display that provides a cohesive, easy-to-read view of flight information. The paper deals with the problems that arise in the development of a certified visualization system for the pilot's display, designed to work on civil aircraft under the Russian real-time operating system JetOS. It is shown that we have successfully solved the problems of multi-window visualization of many aviation applications with an acceptable speed on a promising aviation computing platform. The article lists several algorithmic solutions that allow achieving the required rendering speed, as well as outlines ways for further work.

In this paper, we propose an approach for calculating global illumination by the radiosity method that allows achieving high performance for a large number of patches involved in the calculations. Two key ideas of the proposed approach are the creation of proxy geometry based on axis-aligned virtual patches and the method of temporal radiosity, which uses the interpretation of form-factors in terms of probabilities. The combination of the proposed methods makes it possible to obtain a rapid update of global illumination for light sources changing their positions with image quality close to the naive radiosity method.

Virtual sensing technology uses mathematical calculations instead of natural measurements when the latter are too difficult or expensive. Nowadays, application of virtual light sensing technology becomes almost mandatory for daylight analysis at the stage of architectural project development. Daylight Autonomy metrics should be calculated multiple times during the project. A properly designed building can reduce the necessity of artificial lighting, thus saving energy. There are two main daylight performance metrics: Spatial Daylight Autonomy (sDA) and Annual Sunlight Exposure (ASE). To obtain their values, we have to simulate global illumination for every hour of the year. A light simulation method should therefore be as efficient as possible for processing complex building models. In this paper we present a method for fast calculation of Daylight Autonomy metrics, allowing them to be calculated within a reasonable timescale. We compared our method with straightforward calculations and other existing solutions. This comparison demonstrates good agreement; this proves sufficient accuracy and higher efficiency of the method. Our method also contains an original algorithm for the automatic setting of the sensing area. The sDA metric is calculated considering blinds control, which should open or close them depending on overexposure to direct sunlight. Thus, we developed an optimization procedure to determine the blinds configuration at any time.

The results of numerical simulation of a vertical-axis wind turbine (VAWT) based on the solution of three-dimensional Reynolds-averaged Navier–Stokes equations with the Spalart–Allmaras turbulence model are presented. The results of parametric calculations of a viscous compressible flow under conditions simulating urban infrastructure for a helicoid-type wind turbine with three spirally twisted blades are presented.

The work provides an overview of methods aimed to the reconstruction of Bidirectional Scattering Distribution Function (BSDF) for rough surfaces.The elements with rough surfaces are permanently present in our life and widely used in modern optical devices, for example, in light guiding plates for display illuminating systems, car dashboards, or luminaires.Light scattering by rough surface is an important component in the visual appearance of many materials including water, glass, skin, etc.The problem of the rough surface visualization is complex and contains many different aspects.Accordingly there are many techniques to provide their realistic rendering.In many lighting simulation and optical design tasks it is sufficient and more effective to replace real geometry of rough surface by a surface optical characteristics expressed via BSDF.So, accurate reconstruction of scattering properties of rough surfaces is a significant factor in visualizations tasks and generation of photorealistic images.In some cases, BSDF can be just measured.However, in many cases direct BSDF measurements are impossible if, for example, it is required to define BSDF inside of the material and neither a measuring device detector nor a light source can be placed inside the material.So this results in the development of many approaches for BSDF reconstruction.It started in the end of the last century with the development of many analytical methods based on microfacet models of rough surface such as the Phong, the Ward reflection, the Cook-Torrance models.Nowadays many direct numerical methods of BSDF reconstruction appear, for example, methods based on normals and heights distribution.As a rule, these methods use ray tracing to calculate BSDF.Sizes of microroughness can be small, sufficient to raise a problem which optics wave or ray is more appropriate here.To answer this and other questions related to BSDF reconstruction, an investigation of well-known and effective reconstruct methods was conducted.This paper also presents the study results for eight real samples with different profile parameters of rough surface.The verification is based on numerical comparison with real measured data and visual comparison of images generated using different reconstructed BSDF.Finally, the general recommendations are presented about what methods and for what applications are more appropriate.

The main purpose of this investigation is to use information from previous frames for partial synthesis of the new one. Algorithms utilizing already calculated information have become an integral part of many interactive applications in recent years, and using of such methods allows reduces the computational load on the GPU without significantly degrading the image quality. Despite its popularity, there are still serious unresolved issues that affect image quality in many scenarios. Especially acute are the issues of artifacts and frame distortions. This paper discuss ways of counting and using information from previous frames used in image rendering applications, and proposes a new algorithm which combine best traits of the previous ones.

Software used in avionics must comply with strict aviation standards. The aviation standard ARINC 661 defines the interfaces and a number of specific requirements for the information display system in the cockpit. Due to these requirements, the task of ensuring an acceptable rendering speed on the promising i.MX6 platform with reduced power consumption arises. The development of the OpenGL SC (Safety Critical) graphics library that works in the JetOS real-time avionics operating system and accomplishes this task is considered. It displays information generated in accordance with the ARINC 661 standard using the hardware support of the Vivante GPU. An efficient rendering acceleration approach is proposed and implemented. First of all, calls to OpenGL functions were optimized in the display system that works in accordance with the standard. However, such a modification incurs significant expenses for certifying the entire display system. For this reason, optimization was implemented in a special intermediate module. The proposed approach makes it possible to achieve a rendering speed that is acceptable for the pilot's display in the cockpit.

Nowadays the daylight analysis is widely used by architects because it influences on the room usage and comfortability. The natural daylight has a lot of benefits. There are many issues related to the daylight which affects the function and perception of the illuminated area. The issues are direct sunlight and discomfort glare, illumination level and distribution on the working plane. The correct building design that takes into account the daylight analysis aspects leads to decrease of the artificial light usage and energy saving. So simulation of the metrics used in daylight analysis is an important task. The paper considers the simulation of two main standard characteristics widely used in the modern daylight estimation: spatial Daylight Autonomy (sDA) and Annual Sunlight Exposure (ASE). The calculation of both metrics is regulated by the Illuminating Engineering Society (IES) standards. The metric calculation provides daylight illumination data for whole year with one hour step. This means that several thousand simulations have to be run. So effective light simulation methods must be elaborated and used here. The paper presents the methods which deal with rather complex and precise building models under daylight illumination and realistic environment. Spatial Daylight Autonomy metric is calculated considering blinds control that opens or closes them depending on the over-exposure by the direct sunlight. Thus, simulation of sDA involves an optimization procedure defining blinds configuration at each hourly moment. The sDA and ASE calculations for architecture scene are provided as examples.

The paper continues a series of publications of the authors' research materials in the field of developing an approach to dynamic planning and control of a generalized computational experiment based on visualization methods and visual analytics.A generalized computational experiment involves multiple solution of a numerical simulation problem for different sets of values of model defining parameters, which makes it possible to obtain a solution immediately for a certain class of mathematical modeling problems specified in a multidimensional parameter space.The paper considers an extension of the existing authors' approach to analysis of a generalized computational experiment state using visual maps, based on visualization metaphors that can display not only individual images but also their relationships.A method is proposed for constructing visual maps of a generalized computational experiment focused on visualizing relationships between single computational experiments in three-dimensional space.The method is based on the mechanism of formalization of the relationships between single computational experiments, as well as the concept of a graph model visualization metaphor that defines a visual map prototype.A description is given of a software system for constructing and analyzing three-dimensional visual maps of a generalized computational experiment.The paper also considers examples of its application in estimating the accuracy of numerical models of the OpenFOAM software platform for a three-dimensional problem of inviscid flow around a cone.

The rapid development of machine learning technologies in recent years has led to the emergence of CNN-based sensors or ML-enabled smart sensor systems, which are intensively used in medical analytics, unmanned driving of cars, Earth sensing, etc. In practice, the accuracy of CNN-based sensors is highly dependent on the quality of the training datasets. The preparation of such datasets faces two fundamental challenges: data quantity and data quality. In this paper, we propose an approach aimed to solve both of these problems and investigate its efficiency. Our solution improves training datasets and validates it in several different applications: object classification and detection, depth buffer reconstruction, panoptic segmentation. We present a pipeline for image dataset augmentation by synthesis with computer graphics and generative neural networks approaches. Our solution is well-controlled and allows us to generate datasets in a reproducible manner with the desired distribution of features which is essential to conduct specific experiments in computer vision. We developed a content creation pipeline targeted to create realistic image sequences with highly variable content. Our technique allows rendering of a single 3D object or 3D scene in a variety of ways, including changing of geometry, materials and lighting. By using synthetic data in training, we have improved the accuracy of CNN-based sensors compared to using only real-life data.