Articles
Bridging digital shadows with physiological models and field applications
Article number
1433_30
Pages
235 – 242
Language
English
Abstract
Phenotyping in horticulture, which involves the quantitative and qualitative assessment of plant traits, plays a critical role in crop monitoring and precision agriculture.
Traditionally, manual phenotyping has been labour-intensive and prone to errors.
However, advancements in digital technologies, such as LiDAR, RGB-D cameras, and AI, have enhanced the accuracy and efficiency of plant trait analysis.
These innovations enable the creation of digital twins, deep learning models, and real-time plant monitoring systems.
Recent developments focus on in situ data collection and fruit detection using multi-sensor systems.
LiDAR-based systems have shown promise in estimating plant structure and fruit size while overcoming challenges like lighting variability.
Combining various sensors with AI facilitates precise applications, such as pesticide spraying, selective harvesting, and weed management, significantly reducing chemical use and enhancing efficiency.
Future research aims to improve the integration of phenotypic data into physiological models to improve site-specific management by means of real-time applications.
Key focus areas include spatio-temporal monitoring of structural parameters, increasing model resolution, and leveraging transfer learning across sensors.
The synergy between digital tools and farmers will further advance site-specific applications, improving the precision of variable rate tasks such as irrigation, pruning, and harvesting.
As robotics, real-time applications, and AI continue to evolve, they hold the potential to revolutionize horticulture through enhanced decision-making, optimized resource use, and reduced environmental impact.
Traditionally, manual phenotyping has been labour-intensive and prone to errors.
However, advancements in digital technologies, such as LiDAR, RGB-D cameras, and AI, have enhanced the accuracy and efficiency of plant trait analysis.
These innovations enable the creation of digital twins, deep learning models, and real-time plant monitoring systems.
Recent developments focus on in situ data collection and fruit detection using multi-sensor systems.
LiDAR-based systems have shown promise in estimating plant structure and fruit size while overcoming challenges like lighting variability.
Combining various sensors with AI facilitates precise applications, such as pesticide spraying, selective harvesting, and weed management, significantly reducing chemical use and enhancing efficiency.
Future research aims to improve the integration of phenotypic data into physiological models to improve site-specific management by means of real-time applications.
Key focus areas include spatio-temporal monitoring of structural parameters, increasing model resolution, and leveraging transfer learning across sensors.
The synergy between digital tools and farmers will further advance site-specific applications, improving the precision of variable rate tasks such as irrigation, pruning, and harvesting.
As robotics, real-time applications, and AI continue to evolve, they hold the potential to revolutionize horticulture through enhanced decision-making, optimized resource use, and reduced environmental impact.
Authors
N. Tsoulias
Keywords
optical sensors, real-time application, robotics, proximal sensing, precision horticulture
Online Articles (41)