A weed control approach in Christmas tree production based on tree crown detection using remote sensing and deep learning

Leif O. Harders; Thorsten Ufer; Andreas Wrede; Stephan Hussmann; Eberhard Hartung

doi:10.1117/12.3016507

18 June 2024 A weed control approach in Christmas tree production based on tree crown detection using remote sensing and deep learning

Leif O. Harders, Thorsten Ufer, Andreas Wrede, Stephan Hussmann, Eberhard Hartung

Author Affiliations +

Proceedings Volume 12998, Optics, Photonics, and Digital Technologies for Imaging Applications VIII; 129981A (2024) https://doi.org/10.1117/12.3016507
Event: SPIE Photonics Europe, 2024, Strasbourg, France

Conference Poster

Abstract

Advanced information technologies, such as the fifth-generation mobile networks (5G), unmanned aerial vehicles (UAV), and artificial intelligence (AI), build the foundation for integrated autonomous systems in smart agriculture that contribute to the sustainable transformation and optimization of agricultural processes, such as weed control. Weed control is a particular challenge for specialty crops, as it is labor-intensive, and the widely used chemical weed control is increasingly subject to legal restrictions. A new Regulation on the Sustainable Use of Plant Protection Products was adopted by the European Commission in 2022 that targets to reduce the use of chemical pesticides by 50% by 2030. This paper proposes a sustainable weed control approach based on tree crown detection using remote sensing data and evaluates the state-of-the-art deep learning architectures YOLOv4-tiny, YOLOv4-tiny-3l, and YOLOv7-tiny using a five-fold cross-validation with image inputs sizes of 832 x 832 and 1152 x 1152 pixels. The deep learning architectures are trained and evaluated with a custom dataset of 63 individual UAV-based images with 1,380 transplanted three-year-old Christmas trees under normal production conditions. The selected deep learning models achieve an average precision (AP) at an intersection over union (IOU) threshold of 0.5 above 97%, showing that all selected architectures are suited for this specific application. Neither a significant interaction effect nor significant main effects of the deep learning architecture or the image input size could be observed on the AP.

(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Leif O. Harders, Thorsten Ufer, Andreas Wrede, Stephan Hussmann, and Eberhard Hartung "A weed control approach in Christmas tree production based on tree crown detection using remote sensing and deep learning", Proc. SPIE 12998, Optics, Photonics, and Digital Technologies for Imaging Applications VIII, 129981A (18 June 2024); https://doi.org/10.1117/12.3016507

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

;

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE