Исследовательская база

Высокоточный пчелиный улей с открытым исходным кодом для долгосрочного наблюдения и исследования медоносных пчёл

Computer vision can automate observation-hive research, but differences in hive geometry and imaging conditions make experiments and trained models difficult to reproduce. This paper presents an adaptable open-hardware observation hive that standardizes physical and optical properties for long-term honeybee monitoring. The authors demonstrate the platform with an open-source system for focused tracking of an individual bee and publish the supporting hardware designs, software, and supplementary resources.

Сведения о публикации

Авторы
Jiří Ulrich, Martin Stefanec, Tomáš Rouček, Laurenz Alexander Fedotoff, Harald Pascher, Zdeněk Rozsypálek, Thomas Schmickl, Tomáš Krajník
Организации
🇨🇿 Czech Technical University in Prague🇦🇹 University of Graz
Год
2026
Тип
Journal

Relevancy to Gratheon

The paper addresses an important deployment problem for Gratheon's computer-vision systems: models trained with one hive and camera geometry often do not transfer cleanly to another. Its standardized, washable observation-hive frame and fixed optical configuration provide practical references for repeatable camera placement, illumination, calibration, and multi-site data collection.

The open designs are also directly useful for the hive-scanner roadmap. They can inform prototype fixtures and controlled validation environments where Gratheon can compare tracking and frame-analysis models without introducing uncontrolled changes in image geometry.