2687-0517 19 3 2026 1-147

RAR RUS 7-17 0000-0003-0726-6613 Peter the Great St. Petersburg Polytechnic University Pavlov Vitalii pavlov_va@spbstu.ru

St. Petersburg, Russian Federation

0009-0008-1065-6798 Tsytsulin Aleksandr tsysulin@niitv.ru 0000-0001-5282-2506 Savischenko Nikolay Russian-Armenian University Avetisyan Vaan avahan@mail.ru

Yerevan, Republic of Armenia

Identification of spacecraft at close range using depth maps The article addresses the problem of automatic detection, segmentation, and classification of spacecraft using depth maps. The study was based on synthesized semi-natural data, which made it possible to simulate various orbital imaging conditions and to provide accurate ground-truth annotations for the quantitative evaluation of algorithms. Due to the instability of texture and color features under real space conditions, the geometric structure of objects is proposed as a key invariant feature. To solve the detection problem, a multi-stage algorithm has been developed, which includes adaptive histogram equalization, local threshold segmentation, morphological processing, and filtering of false contours based on geometric criteria. To classify spacecraft, deep neural networks of the ResNet family were used in two configurations: based on optical images alone and based on combined data (image + depth map). Experiments on a synthetic dataset of 19440 images showed that the proposed algorithm significantly outperforms the SAM 3 model in terms of detection accuracy (Precision 0.917 vs 0.171), ensuring a low level of false positives. The multimodal approach to classification achieved an accuracy of 99.9% (ResNet-152 model), confirming the effectiveness of using depth maps for reliable recognition of spacecraft under near-real space imaging conditions. 10.18721/JCSTCS.19301 004.932.72 spacecraft synthetic data object detection computer vision depth maps https://infocom.spbstu.ru/article/2026.90.1/

RAR RUS 18-28 0009-0009-6909-9399 National Research University “Moscow Power Engineering Institute”; Moscow Technical University of Communications and Informatics Agamirov Levon itno_agamirov@mail.ru

Moscow, Russian Federation

0000-0001-9181-7726 Moscow Technical University of Communications and Informatics; Moscow Aviation Institute (National Research University) Agamirov Vladimir avhere@yandex.ru

Moscow, Russian Federation

0009-0009-6284-515X Slyadnikov Pavel p.e.slyadnikov@mtuci.ru 0000-0002-2851-8472 Moscow Technical University of Communications and Informatics Toutova Natalia e-natasha@mail.ru

Moscow, Russian Federation

0009-0008-7929-5492 Moscow Technical University of Communications and Informatics Khaush Anton anton.khaush@gmail.com

Moscow, Russian Federation

Design of an automated freight transportation monitoring system based on internet of things networks using blockchain technology The paper proposes an architecture for an automated freight transportation monitoring system based on blockchain technology. When integrating standard blockchain solutions with Internet of Things (IoT) devices into a unified techno-logical system, it is necessary to minimize direct requests to the main registry (root blockchain). This requirement remains essential even when participants of the distributed network are granted permissions to modify smart contract parameters, thereby reducing the load on the root registry while ensuring a high level of security during data transmission: from sensors to the mobile node and from the mobile node to the root system. The paper presents a comparative analysis of distributed ledger platforms, taking into account regional specifics and the possibility of integrating certified domestic cryptographic information protection tools. A prototype of an end-to-end freight transportation monitoring system has been developed and implemented, providing continuous tracking of cargo conditions sensitive to environmental factors. Simulation results confirm the effectiveness of using smart contracts for automating control procedures and ensuring the immutability of logistics operation data. The proposed architecture enables the development of scalable and secure industrial monitoring networks for logistics companies. 10.18721/JCSTCS.19302 004.8 Internet of Things blockchain smart contract cargo transportation logistics mobile node information security https://infocom.spbstu.ru/article/2026.90.2/

RAR RUS 29-45 0009-0005-6368-0001 Goriunov Maksim goryunov_myu@spbstu.ru Peter the Great St. Petersburg Polytechnic University Konstantinov Andrei andrue.konst@gmail.com

St. Petersburg, Russian Federation

Novikov Fedor Kernel Regression with Tree-Induced Metrics This paper presents a family of kernel models based on metrics derived from the structure of decision trees and random forests. Unlike the standard random-forest proximity, which only checks whether two observations fall into the same leaf, the proposed metrics also use information about decision paths, the depth of their divergence, and the depth of their lowest common ancestor. A weighted variant with trainable depth-dependent parameters is also introduced. The tree-based metrics are aggregated over a forest and used to construct exponential kernels. They can also be combined with the Euclidean distance in the original feature space. We prove that the resulting kernels are positive semidefinite and can therefore be used in standard kernel methods, including Nadaraya–Watson regression and kernel ridge regression. The proposed models are evaluated on classification datasets and in a controlled source-to-target regression experiment with a limited number of labeled target observations. The results show that information about the internal structure of decision trees can be useful beyond simple leaf coincidence. Several proposed variants improve upon conventional geometric kernels and the standard random-forest proximity kernel, while the best-performing metric depends on the prediction method and the dataset. 10.18721/JCSTCS.19303 004.852 kernel regression decision trees tree-based metrics random forests Nadaraya–Watson estimator kernel ridge regression https://infocom.spbstu.ru/article/2026.90.3/

RAR RUS 46-56 0000-0001-8553-640X Kudayeva Fatimat kfatimat@yandex.ru 0009-0009-4666-3356 Bitsuev Timur timur_kbsu@mail.ru Hybrid intelligent system for stratifying the risk of coronary heart disease The article proposes a hybrid intelligent system for stratifying the risk of coronary heart disease (CHD). It includes ensemble forecasting, natural language processing, and explainable AI technologies. It is demonstrated that the integration of structured clinical features and unstructured text data improves the accuracy of CHD risk assessment compared to the traditional SCORE2 scales. On test datasets, high values of AUC-ROC and F1-score were achieved, while the interpretability of the predictions is ensured by constructing waterfall plots that reflect the contribution of key risk factors. The developed software package will be useful in cardiology for supporting clinical decision-making. 10.18721/JCSTCS.19304 004 hybrid intelligent system ensemble machine learning cardiovascular outcome predic-tion natural language processing artificial intelligence SHAP analysis clinical decision support https://infocom.spbstu.ru/article/2026.90.4/