Олимпиада Read more
Quotation gallery
To teach is to learn twice.
Joseph Joubert
Faculty News
Пожароопасный сезон в лесах СКО в самом разгаре. Правила обращения с огнем и любовь к природе – первоосновы воспитания наших маленьких сограждан. В ра Read more
On April 17, 2026, as part of the Week of Science, the role-playing game «Court in a Criminal Case» was held. The event was held as an open criminal t Read more
On April 17, 2026, a solemn ceremony was held at Kozybaev University to award personalized scholarships dedicated to the legacy of Alexander Sergeyevi Read more
On April 16, 2026, at the Faculty of History, Economics, and Law, a solemn ceremony was held to equip the certified 1C-Accounting center with modern c Read more
On April 10, 2026, the Faculty of History, Economics, and Law hosted a session titled «Legal and Socio-Economic Aspects of Society's Development in th Read more
Гранты Read more
The M. Kozybayev North Kazakhstan University hosted the International Scientific and Practical Conference «Kozybayev Readings 2026», dedicated to the Read more
On April 8, 2026, the educational programs developed by the Academy of the International Financial Center «Astana» were officially transferred to the Read more
7 апреля 2026 года медицинский факультет НАО «Северо-Казахстанский университет им.М. Козыбаева» в рамках празднования Международного дня здоровья о Read more
Development of a Neural Network Model for UAV Recognition through an Optical-Electronic Channel Integrated into the Data Fusion System
This research is funded by the Science Committee of the Ministry of Science and Higher Education of the Republic of Kazakhstan (grant No. AR19679009).
Project Manager: Kurmashev I.G., Candidate of Technical Sciences.
Project Performers: Kurmashev I.G., Candidate of Technical Sciences, Serbin Vasily Valeryevich, Candidate of Technical Sciences, Arrichiello Filippo, Doctor of Technical Sciences, Semenyuk V.V., Master, Alyoshin D.V., Master, Kryuchkov V.N., Master, Kurmasheva L.B., Master.
Implementation period: 2023 – 2025.
Project goal: To develop a software model for UAV recognition based on neural networks, adapted to the "FMCW radar + video surveillance" platform, performing high-quality and highly accurate recognition, classification, and differentiation of these objects from birds by analyzing the optical channel and micro-Doppler characteristics of the target. Expected outcomes: Development of a software model for UAV recognition based on algorithms of two types of neural networks, adapted to the optical and radar channel of the "FMCW radar + video surveillance" system. Project description: The project idea is to create a neural network software model, one part of which is designed for UAV recognition through radar imaging of micro-Doppler signatures, ensuring more accurate classification of drones and birds. The second segment of the software model defines a neural network application for UAV recognition through video data and photo images of objects in the airspace (quadcopters, "flying wing" UAVs, birds, etc.). The unique feature of this development lies in its adaptation to an Anti-Drone radar system with a software-hardware platform based on "Radar + Optical Channel" as an automation element for UAV recognition using two detection channels. The performance and effectiveness of the developed software model depend on the characteristics of the radar system and optical camera, making the selection and justification of the radar model and surveillance tool a key task. Additionally, the project will present the mathematical characteristics of radar signal reflection from targets with vibration sources, determining Doppler indicators for recognizing flying objects (UAVs and birds). The structural description of the Data Fusion system, into which the developed software model is integrated, along with the characteristics of the neural network algorithms serving as the basis for classification and recognition software within the research framework, will also be provided.