Computer Vision Based Recognition Algorithm for Sea Buckthorn Harvesting Robot

2024 IEEE 65th International Scientific Conference on Power and Electrical Engineering of Riga Technical University (RTUCON 2024): Conference Proceedings 2024
Jurijs Timofejevs, Mihails Gorobecs, Andrejs Potapovs

Sea buckthorn harvesting presents challenges in labour requirements and costs, urging the need for technological advancements. This paper discusses the applicability of adaptive systems, computer vision, and robotics in agricultural tasks, focusing on the development of an autonomous computer vision-based sea buckthorn harvesting robot. The paper outlines the overall architecture of the robot and presents results demonstrating its detecting ability. Additionally, a branch direction detection algorithm is described. The research shows satisfactory precision levels for detection on known samples, although precision for new samples requires improvement in future research steps. The outlined conclusions highlight the achieved milestones in the development of the sea buckthorn harvesting robot and emphasise areas for further refinement and enhancement.

Atslēgas vārdi
Training , Electrical engineering , Computer vision , Image recognition , Costs , Adaptive systems , Computational modeling , Focusing , Computer architecture , Detection algorithms, Harvesting , Deep neural network , machine learning
DOI
10.1109/RTUCON62997.2024.10830877
Hipersaite
https://ieeexplore.ieee.org/document/10830877

Timofejevs, J., Gorobecs, M., Potapovs, A. Computer Vision Based Recognition Algorithm for Sea Buckthorn Harvesting Robot. No: 2024 IEEE 65th International Scientific Conference on Power and Electrical Engineering of Riga Technical University (RTUCON 2024): Conference Proceedings, Latvija, Riga, 10.-12. oktobris, 2024. Piscataway: IEEE, 2024, 1.-5.lpp. ISBN 979-8-3503-6578-8. e-ISBN 979-8-3503-6577-1. ISSN 2996-1033. e-ISSN 2768-3338. Pieejams: doi:10.1109/RTUCON62997.2024.10830877

Publikācijas valoda
English (en)