ADAPTIVE TERRAIN-REFERENCED VISION SYSTEM FOR AUTONOMOUS MULTICOPTER NAVIGATION
Problems of Information Technology · 17.07.2026 · 17 (2) · 45-55
DOI: 10.25045/jpit.v17.i2.05 ↗Kurz gefasst
Fachartikel in der Zeitschrift Problems of Information Technology, online erschienen am 17.07.2026. Verfasst von Vagif Aliyev.
Abstract (Originalsprache)
This paper proposes an onboard computer vision system with adaptive selection of recognition methods for autonomous multicopter navigation in Global Navigation Satellite Systems (GNSS)-denied environments. The approach represents terrain images as onedimensional structural signals obtained through preprocessing, segmentation, grayscale conversion, and linearization. Recognition is performed using spectral, wavelet, and dynamic signal matching methods. An adaptive mechanism selects the most appropriate recognition method based on adequacy criteria describing the uniformity, symmetry, convergence, and sensitivity of the recognition process. The selected or combined methods are used for terrain recognition, checkpoint localization, trajectory reconstruction, and flight correction within a unified closed-loop navigation framework. The proposed architecture enables reliable autonomous navigation under spatial distortions, changing observation conditions, and limited onboard computational resources. Experimental results demonstrate that adaptive selection and weighted combination of recognition methods improve the robustness and accuracy of terrain image recognition for navigation in GNSSdenied conditions.
Zitieren
APA
Aliyev, V. (2026). ADAPTIVE TERRAIN-REFERENCED VISION SYSTEM FOR AUTONOMOUS MULTICOPTER NAVIGATION. Problems of Information Technology, 17(2), 45-55. https://doi.org/10.25045/jpit.v17.i2.05
Vancouver
Aliyev V. ADAPTIVE TERRAIN-REFERENCED VISION SYSTEM FOR AUTONOMOUS MULTICOPTER NAVIGATION. Problems of Information Technology. 2026;17(2):45-55. doi:10.25045/jpit.v17.i2.05
BibTeX
@article{Aliyev2026,
author = {Aliyev, Vagif},
title = {ADAPTIVE TERRAIN-REFERENCED VISION SYSTEM FOR AUTONOMOUS MULTICOPTER NAVIGATION},
journal = {Problems of Information Technology},
year = {2026},
volume = {17},
number = {2},
pages = {45--55},
doi = {10.25045/jpit.v17.i2.05}
}
EUVIMED