Drone-based phenotyping using unmanned aerial vehicles (UAVs) has emerged as a revolutionary approach for

high-throughput, precise, and scalable measurement of plant traits critical to crop improvement. This technology

assessment, yield prediction, and accelerating breeding cycles by facilitating objective, rapid selection of

superior genotypes across multiple crop species. Despite its transformative potential, challenges remain in

standardizing protocols, managing large-scale complex datasets, integrating phenotypic with genomic and

environmental data, and providing training resources for widespread adoption. Ongoing advancements in sensor

technology, data analytics, open-source tools, and capacity building are poised to cement drone-based

phenotyping as a cornerstone technology for sustainable, climate-resilient crop breeding and global food

security.

Keywords: UAV-based phenotyping, High-throughput phenotyping; Drone remote sensing; Precision

agriculture, Crop improvement; Plant breeding; Stress detection; Climate-resilient agriculture.

evaluations often suffer from inconsistency and are unsuitable for screening large populations across diverse

environmental conditions.

The integration of unmanned aerial vehicles (UAVs), commonly referred to as drones, equipped with

multispectral, hyperspectral, thermal, and RGB imaging sensors, has revolutionized plant phenotyping. This

technological advancement facilitates high-throughput phenotyping (HTP) by enabling rapid, accurate, and non-

destructive trait data acquisition at different growth stages and across expansive field trials (Bhandari et al., 2023;

Yang et al., 2022). UAV-based platforms overcome major limitations of traditional phenotyping by providing

scalable data collection with high spatiotemporal resolution, thus capturing dynamic phenotypic traits reflective

of genotype-environment interactions under real field conditions (Patil et al., 2024).

This phenomics revolution is synergistically linked with genomics, where precise phenotypic datasets generated

This review synthesizes recent advances in drone-based phenotyping technologies, their methodological

frameworks, data processing techniques, including artificial intelligence applications, and the practical

implications for contemporary crop improvement. It further discusses current challenges, including data

standardization and accessibility, and outlines future research directions critical for maximizing the impact of

UAV phenotyping in sustainable agriculture.

Unmanned aerial vehicles (UAVs), equipped with advanced imaging sensors such as red-green-blue (RGB)

cameras, multispectral, hyperspectral, and thermal sensors, have emerged as powerful tools for high-throughput

plant phenotyping in field-based crop improvement programs (Patil et al., 2024; Yang et al., 2022). These drones

The major advantage of UAV platforms over traditional satellite remote sensing lies in their superior spatial and

temporal resolution. Satellite imagery typically suffers from infrequent revisit times and insufficient resolution

to accurately capture variability within small breeding plots, limiting its utility in high-precision breeding efforts

and disease progression (Patil et al., 2024). This trait-by-trait temporal monitoring provides breeders with

detailed spatiotemporal phenotypic information that correlates more directly with genotype performance in target

environments.

associated with manual phenotyping (Jangra et al., 2021). Additionally, the cost-effectiveness of UAV operations

enables the screening of thousands of genotypes rapidly across multiple environments, providing breeders with

powerful datasets to select superior germplasm (Marcone et al., 2024). The integration of UAV phenomics with

advanced analytics, including machine learning and deep learning, further enhances trait extraction accuracy and

predictive modeling capacity, driving precision breeding (Yang et al., 2022).

Furthermore, the ability of drones to operate across diverse terrains and environmental conditions adds

robustness to phenotyping campaigns, enabling comprehensive assessments of genotype × environment

from large-scale aerial data. The integration of machine learning (ML) and deep learning (DL) algorithms has

significantly accelerated and enhanced the accuracy of phenotypic trait quantification under field conditions

(Alahmad et al., 2025; Volpato et al., 2024).

Convolutional Neural Networks (CNNs) combined with Long Short-Term Memory (LSTM) models represent a

powerful hybrid approach for analyzing time-series UAV imagery. CNNs efficiently extract spatial features from

individual images, while LSTMs model temporal dependencies by processing sequences of extracted features,

capturing dynamic plant development traits such as relative maturity. This approach has been successfully

applied to dry beans, allowing high-accuracy estimation of phenological stages by leveraging multispectral UAV

data captured across the growing season (Volpato et al., 2024; Zhang et al., 2025).

For early-season applications like stand count estimation, Faster Region-based CNN (Faster R-CNN) algorithms

Algorithms perform ground classification using morphology- and curvature-based filters to differentiate

vegetation from terrain, which refines Digital Terrain Models (DTM) needed to calibrate height estimates. These

technological advances provide non-contact, repeatable, and high-throughput structural trait assessments

imperative for assessing plant vigor, lodging risk, and yield potential (Pun Magar et al., 2025; Zhang et al., 2003).

Additionally, integration of environmental covariates such as growing degree days (GDD) and weather variables

enhances the ecological validity and predictive accuracy of phenotyping models by accounting for temporal and

spatial variability in crop development (Alahmad et al., 2025). Open-source pipelines such as MatchPlant offer

user-friendly, modular frameworks that integrate UAV image processing, annotation, CNN-based detection, and

spatial trait extraction, democratizing drone phenotyping for researchers and breeders worldwide (Sangjan et al.,

2025).

Collectively, these technological and analytical innovations are transforming phenotyping from a bottleneck to

Drone-based phenotyping has been widely adopted across diverse crop species, proving invaluable for improving

Quantitative trait measurement is another critical application, where drones provide non-destructive, high-

throughput estimation of important agronomic traits such as plant height, biomass, leaf area index, and nitrogen

content. These traits are essential for understanding growth dynamics and making selections in breeding

programs (Patil et al., 2024). For instance, Digital Surface Models (DSMs) derived from UAV imagery enable

efficient plant height measurements across large field trials, offering enhanced precision compared to manual

methods (Volpato et al., 2024; Pun Magar et al., 2025).

Yield prediction models have also been dramatically improved using temporal UAV data. By integrating

vegetation indices calculated from RGB and multispectral images taken at critical growth stages with machine

learning techniques, researchers have enhanced the accuracy and robustness of predictions for final grain yield

Alahmad et al., 2025).

of UAV phenotyping to accelerate selection, reduce costs, and improve decision-making in modern breeding

pipelines (Volpato et al., 2024; Yang et al., 2022).

Overall, drone-based phenotyping constitutes a powerful and versatile toolset that integrates high-resolution

aerial data, advanced analytics, and environmental parameters to support precision crop improvement targeted

at increasing productivity, resilience, and sustainability.

comparability and reproducibility across studies and breeding programs. Establishing universal standards for

UAV operations, sensor calibration, data acquisition, and trait extraction is imperative to generate high-quality,

interoperable phenotypic datasets (Kefauver, Araus, & Buchaillot, 2019; Guo et al., 2021).

Handling large and complex datasets generated by high-resolution UAV imaging platforms presents another

Integration of UAV-generated phenotypic data with genomic and environmental datasets is a vital frontier for

precision breeding. Combining high-throughput phenotyping with high-density genotyping and environmental

metadata allows for the comprehensive dissection of genotype × environment interactions and precise genomic

selection. However, integrating heterogeneous data types requires interoperable databases, sophisticated

statistical models, and accessible software platforms tailored for breeders’ use. This necessitates

multidisciplinary collaboration bridging plant science, genomics, bioinformatics, and data science (Marsh et al.,

2021; Sinha et al., 2023).

Another limiting factor is the shortage of accessible training resources, user-friendly tools, and capacity-building

initiatives to facilitate widespread UAV phenotyping adoption in developing countries and breeding

communities with limited technical expertise. Developing comprehensive educational programs, hands-on

workshops, and open-source software with intuitive graphical interfaces will be crucial to democratize drone

phenotyping technology and maximize its global impact (Parker et al., 2022; Excellenceinbreeding.org, 2019).

Future directions should also focus on policy frameworks for UAV use and ethical considerations regarding data

privacy and ownership. Advancements in sensor payloads, battery life, and autonomous flight capabilities will

Drone-based phenotyping has emerged as a transformative tool in crop improvement by overcoming the

limitations of traditional manual assessment methods. It enables rapid, accurate, and high-throughput

measurement of plant traits directly under field conditions. Using advanced sensors such as RGB, multispectral,

hyperspectral, and thermal cameras, drones can capture critical phenotypic traits, including plant height,

biomass, canopy temperature, growth stages, and disease symptoms with high precision. The integration of

https://airborninsight.com.au/drone-based-phenotyping/

10. ICRISAT. (2020). Applications of UAVs: Image-based plant phenotyping. International Crops Research

Institute for the Semi-Arid Tropics.