Section 2 - Track 1

On this page, talks in the format of posters are posted. They can be viewed during the whole time of the Conference. To send your question to the authors, please follow to Miro and write your question in the comments. You can also use the feedback form on our website. Authors will answer questions using Miro comments or via email. To get manual go to Instructions (eng / rus).

Anatoly Novikov and Anton Pronkin
Method and program for detecting borders of brightness difference
A description is given of a new method for detecting borders of brightness difference in images, a computational algorithm, and a description of a computer program. The proposed method belongs to the class of gradient methods, is an analogy and at the same time an alternative to the well-known Canny method. It was created for use in vision systems of aircraft and robotic systems. The main purpose is to identify the contours of objects of constant presence on the underlying surface of the Earth. In contrast to the Canny method, the partial derivatives are estimated in it using the line mask obtained by the least squares method, assuming an adequate description of the image slice along a row (column) on a finite interval by a linear function. Two versions of the program implementing the new method are presented. One – for scientific research – with manual selection of threshold values for the histogram. The second option with automatic selection of threshold values is for integration into real vision systems. The results of the separation of borders on television images, on x-ray images by the proposed method and, for comparison, by the Canny method are presented.

Aleksandr Borodinov and Vladislav Myasnikov
Pairwise comparisons in finding user preferences
In this paper, we consider the problem of reconstructing functions defined implicitly by the results of pairwise comparisons. In the proposed approach, we apply adaptive transformation to the high-dimensional space. Then we classify the comparisons using linear or non-linear classifiers. In this work, we consider linear regression and random forest as classification algorithms. In experimental analysis, we compare different methods of transformation to the high-dimensional space and investigate the effectiveness of the proposed method.

Natalia Rodionova, Irina Vakhnina and Tatiana Zhelibo
Assessment of vegetation state post-fire dynamics on the territory of Ivano-Arakhleisk natural Park (Zabaikalsky Krai) by Sentinel 1/2 radar and optical data
The result of the analysis of multi-temporal satellite monitoring of the post-fire vegetation state dynamics in the territory of Ivan - Arakhley natural Park (Zabaikalsky Krai) after the fire in 2015 is presented using Sentinel 1 (S1) radar data and Sentinel 2 (S2) optical data. To assess the dynamics of revegetation affected by natural fire, spectral vegetation indices (VI) NDVI, ARVI, NBR, NDMI and radar vegetation index RVI are used. A positive trend has been revealed in the restoration of vegetation in the test areas of the natural Park by both optical and radar indices.

Alina Bavrina, Ludmila Kavelenova, Oksana Kuzovenko and Nataly Prokhorova
Detection and age estimation of burned areas of natural grassy communities in the Samara region using Sentinel-2 data
An analysis of the environmental and socio-economic aspects associated with the steppe fires - periodically recurring natural and natural-anthropogenic emergencies - shows their high importance as a negative phenomenon for the Russian Federation. The article discusses the possibility of the detection of burned areas and their age estimation based on the calculation of spectral indices between two consecutive Sentinel-2 acquisitions. The study was conducted for the natural grassy communities of the Samara region, in which an increase in fires was observed in 2018. Using up-to-date sources of remote sensing allows to obtain additional data for research and analysis of pyrogenic processes in our region.

Anton Agafonov, Aleksey Maksimov and Aleksandr Borodinov
Performance comparison of GPU parallelization algorithms for the reliable shortest path problem
The paper considers the reliable shortest path problem in a stochastic transportation network that maximizes the probability of arriving at a destination within a predetermined period of time (time budget). Existing algorithms solves this problem with good results in terms of the quality (travel time) of the found route, however, but have high computational complexity, which does not allow using them in practical navigational applications in real time. The paper presents a parallelization strategy for finding a reliable path using the CUDA GPU. Experimental studies conducted on the transport network of the Samara city show that the proposed approach can reduce the computation time by an average of 5 times.

Konstantin Dobratulin, Andrey Gaidel, Irina Aupova, Anna Ivleva, Aleksandr Kapishnikov and Pavel Zelter
The efficiency of deep learning algorithms for detecting anatomical reference points on radiological images of the head profile
In this article we investigate the efficiency of deep learning algorithms in solving the task of detecting anatomical reference points on radiological images of the head in lateral projection using a fully convolutional neural network and a fully convolutional neural network with an extended architecture for biomedical image segmentation - U-Net. A comparison is made for the results of detection anatomical reference points for each of the selected neural network architectures and their comparison with the results obtained when orthodontists detected anatomical reference points. Based on the obtained results, it was concluded that a U-Net neural network allows performing the detection of anatomical reference points more accurately than a fully convolutional neural network. The results of the detection of anatomical reference points by the U-Net neural network are closer to the average results of the detection of reference points by a group of orthodontists

Nina Vinogradova, Andrey Sosnovsky and Natalya Sevostyanova
Automatic recognition of the number of channels in unidentified multispectral data
The work is devoted to the development of a method for identifying unknown parameters of multizone Earth images got from remote sensing systems. The method allows automatically to determine the method of alternating spectral channels and calculate their number for images stored in files of uncompressed formats. A theoretical justification based on a change in the shape of the Fourier spectrum with a change in the method of alternating channels in the data file is presented, features of the shape of the spectrum are revealed that allow reliable identification of the required characteristics. The results of applying the algorithm to real Earth images from space are presented, its applicability limits are indicated, and recommendations are given for choosing specific parameters of the algorithm.

Nina Vinogradova and Leonid Dorosinsky
Research of algorithms for detecting small changes over the data of a radar image of the Earth from space
When solving various problems of space monitoring, the problem arises of determining the presence or absence of small changes in the state of the reflecting surface during the observation session, as well as determining the characteristics of these changes, such as the position, size and average power values ​​of the reflected signals. The article is devoted to the development of a method designed to detect such changes, as well as the study of the effectiveness of this method based on estimates of measurement errors. Concrete recommendations are given for the implementation of these method based on data on real scattering cross-section of the region.

Radik Magdeev, Marat Suetin and Aleksandr Tashlinskii
Improving the efficiency of the method of stochastic gradient identification of objects in binary and grayscale images due to their pre-processing
In this paper, we consider ways to improve the efficiency of the method of stochastic gradient identification of objects for binary and grayscale images due to the methods of image preprocessing. Identification of an object is understood as the recognition of an object on the image with its parameters estimation. Low-pass filtering and image equalization are considered as preliminary processing. The rate of convergence of identification parameters is investigated. The optimal sizes of the Gaussian filter mask for binary and grayscale images were found based on COIL-20 images.

Sergey Zraenko
Increasing the distinctiveness of forest species composition by satellite images
The brightness of reflections from coniferous and foliar vegetation was studied using Landsat–7 images for different seasons of the year. Results were obtained for each of the spectral channels of the ETM + sensor, which allows forming standards for classifying forest vegetation in various phenological phases. To increase the distinctness of plant objects, information about their brightness is combined with data from another spectral channel. As a result, an additional classification feature is formed – the Euclidean distance in the space of spectral brightness. It is shown that the combination of two channels can significantly increase the number of informative classification features when mapping forest vegetation.

Vladislav Batshev, Milana Sharikova, Alexander Machikhin, Sergey Boritko, Vitold Pozhar, Alexey Kozlov and Anton Karandin
A compact acousto-optical module for hyperspectral imaging systems
We present an acousto-optical module with built-in optical channel for tunable spectral image filtration and electronics for adjusting and applying ultrasound signals. Small dimensions and USB interface enable its fast and easy integration into existing imagers. We demonstrate the effectiveness of the proposed design for hyperspectral imaging in the wavelength ranges 450-900 nm and 900-1700 nm. Proposed design may be the basis of hyperspectral imagers for various applications.

Roman Kovalenko, Pavel Smirnov and Radik Ibragimov
Use of stochastic adaptation in block method to estimate deformation field for image sequence
The paper researches the block method based on stochastic adaptation, which is used to estimate the deformation field of the image sequence. The similarity model was selected as the deformation model. The method was implemented for two target functions: the mean square inter-frame difference and the inter-frame correlation coefficient. The result of the proposed method was compared with the Motion Vector Field Adaptive Search Technique. The proposed method has a high noise resistance and allows one to reduce the influence of global inter-frame geometric changes.

Aleksandra Danilenko and Anastasia Guzhenko
Use of convolution networks to solve the problem of detection and recognition of state registration signs of vehicles
The paper presents the implementation of a system for license plate detection and recognition in digital images. A comparative analysis of the detection methods was carried out. It led to the selection of the convolutional neural network for the detection module. The customized network architecture was developed on the basis of the neural network GoogleNet. An overview of approaches for license plate detection and recognition in digital images was made. The best solution for the recognition module was determined. The developed system for license plate detection and recognition can identify license plates that use letters of the Latin alphabet and Arabic digits. Testing of the detection and recognition system of license plate were carried out. It proved the reliability of the system and its ability to perform well, even in conditions that might prevent the detection and recognition of license plates. The accuracy of the implemented system was at least 94%.

Dmitry Gavrilov and Dmitry Lovtsov
Processing of visual information in the automated optoelectronic system of ground-space monitoring
One of the most important tasks of image processing is the search for objects in satellite aerospace images. Most of the known methods are applicable only for detecting large objects. The search for small objects whose linear size does not exceed a dozen pixels is particularly difficult. In addition, aerospace images can contain not only single objects, but also groups of objects of the same type, close to each other, which also makes it difficult to automatically solve the problem of separate localization and classification of each object. The aim of this work is to solve the problem of localization and classification of small single objects and groups of objects of the same type in large format images.

Arseny Golovin, Anatoly Demin and Evgenii Sechak
Landmine detection and minefield mapping with the help of multi-angle long-wave infrared hyperspectral data fused with the 3D terrain reconstruction
The article proposes to use multi-angle hyperspectral long-wave infrared remote sensing together with three-dimensional reconstruction of the area to increase the reliability of detection and reduce the frequency of false alarms when searching for subsurface objects - antipersonnel mines, improvised explosive devices and unexploded ordnance in mountainous and hilly areas, where the use of minesweepers is difficult. Multi-angle remote sensing allows to exclude skipping of objects masked and laid at an angle, and to separate the soil containing anomaly objects from ordinary soil and surface irregularities. The concept of an optical-digital complex for minefield mapping is given, the main basis of which is a hyperspectral device that receives data from two optical channels with divided them on the tens spectral channels in the longwave infrared range. One of optic channel scans the nadir and the second channel scans at an angle to the soil surface. The complex also includes a camera of the visible range, receiving a series of images in different spatial planes for further three-dimensional reconstruction. A method for obtaining and combining segmented hyperspectral data with a reconstructed digital terrain model is described for solving the problems of detection of hidden ground and subsurface objects, reconnaissance and planning of humanitarian demining missions on terrain with different slopes of relief.

Back to Top