Researchers from South China Agricultural University (SCAU) and Guangdong Ocean (GDOU) University built a system that uses MMW radar as its core sensor for fruit orchard canopy recognition.
Unlike other sensors that work of light or sound, this radar emits short-wavelength electromagnetic waves.
This means that it can work will in challenging conditions, such as rain, fog, or bright sunlight, allowing it to accurately detect objects even in bad weather.
According to the study, MMW radar has exhibited “excellent adaptability to environmental and meteorological conditions and anti-rain, fog, and light interference”
It added: “In the research related to orchard canopy recognition and characteristics extraction, millimetre-wave radar has broken through the problems such as low accuracy of ultrasonic detection technology and large influence of machine vision by light.”
However, the study added that this technology was still in its infancy in terms of the identification and characteristic extraction of the canopies of fruit tree orchards.
This addresses a critical need for precision production in agriculture.
The technology can identify the the height, width, and volume of a tree. This information can then be used for precision spraying, which will improve efficiency and reduce waste in the process.
The experiments
The researchers highlighted the importance of accurately identifying fruit tree canopies for more precise farming.
“Fruit tree canopies are an important basis for the intelligent management and mechanised operation of orchards. Fruit orchard canopies’ morphological characteristics can provide scientific guidance for the precise application and fertilisation of orchards, the evaluation of fruit tree yield and quality, precise management of orchards.”
In this study, the MMW radar was put through a series of experiments focusing on two main aspects: fruit orchard canopy recognition and fruit tree canopy characteristic extraction.
The recognition experiments were conducted on a plum orchard within SCAU grounds using E-DBSCAN algorithm
According to the study, it was a small arbor orchard managed by artificial planting with a flat terrain and low weeds.
A total of 31 plum trees were selected for this research.
The technology was found to be highly accurate, achieving an F1 score of 96.7%, which comprised of a precision rate of 93.5% and a recall rate of 95.1%.
The next experiment was conducted on a more controlled environment with seven artificial trees.
It evaluated how well the system could identify characteristics such as height, width, and size under simulated spray conditions.
Comparing between normal conditions, the differences expressed in mean relative errors for height, width, and size were 2.3%, 2.1%, and 4.2% respectively.
The researchers concluded that the E-DBSCAN algorithm could still accurately recognise canopies despite influence from the sprays.
Source: Agriculture 2025, 15, 1342.
Fruit Orchard Canopy Recognition and Extraction of Characteristics Based on Millimeter-Wave Radar
Authors: Yinlong Jiang, Jieli Duan, Yang Li, Jiaxiang Yu, Zhou Yang, and Xing Xu
https://doi.org/10.3390/agriculture15131342
