Методологія ресурсоефективної інтелектуальної візуальної навігації і позиціонування автономних авіаційних платформ

Костянтин Дергачов, Є. М. Овдіюк, Андрій Кисельов et al. · The Scientific Issues of Ternopil Volodymyr Hnatiuk National Pedagogical University Series pedagogy · 2026

Subject of research: Intelligent visual navigation and positioning processes for autonomous aerial platforms under conditions of uncertainty and limited computational resources of onboard systems. The purpose of the study: development of a methodology for resource-efficient intelligent visual navigation and positioning that ensures a coherent integration of mathematical models, algorithmic structures, and resource constraints within a unified approach. Research methods: computer vision methods, intelligent video data processing algorithms, mathematical modeling of visual localization processes, methods for the adaptive generation of algorithmic compositions, as well as simulation and experimental verification on embedded computing platforms. Results: A generalized methodology is proposed, based on the principles of systemic consistency of models and algorithms, adaptive accounting for uncertainty in video data, and the intelligent formation of resource-efficient algorithmic compositions. A mathematical model has been developed that formalizes the navigation process as an optimization problem of a generalized quality functional, considering constraints on resources, reliability, and adaptability. To demonstrate the practical implementation of the methodology, a heterogeneous ensemble tracking system with a voting and correction mechanism was proposed, which confirmed an increase in the stability and efficiency of video data processing under resource-constrained conditions.

Cited by 0 works.

Authors