Advances in Developments and Trends of UAV Technology in the Context of Precision Agriculture

Abstract

Agriculture, as a core pathway for advancing modern agricultural development, emphasizes data-driven perception and intelligent decision-making. Unmanned aerial vehicles (UAVs), with advantages such as high-resolution imaging, flexible deployment, and adaptability to diverse terrains, have become an essential tool in this domain. This Editorial synthesizes the key findings from nine representative studies featured in this Special Issue, focusing on recent advancements in UAV-based remote sensing, flight control, and precision spraying. The results indicate that the integration of multispectral imagery with deep learning models significantly enhances crop identification and parameter inversion accuracy. Flight control performance has been greatly improved through innovations such as free-tail configuration optimization and fuzzy sliding mode composite control, ensuring stable operations in complex environments. In the realm of precision spraying, progress in wind vortex regulation and airflow modeling has led to improved droplet deposition consistency and target accuracy. Overall, UAV technologies demonstrate strong potential for cross-disciplinary integration and scalable application, offering robust support for the intelligent transformation of agricultural production.

1. Introduction

Precision agriculture, as a core direction of modern agricultural development, aims to dynamically perceive crop growth conditions and implement precise regulation through efficient and refined data acquisition and intelligent operation systems. This approach not only enhances yield and quality but also optimizes resource input and minimizes environmental impact. In this process, the deep integration of information technology and intelligent equipment has become a key driving force behind agricultural transformation. As a representative technology in smart agricultural equipment, unmanned aerial vehicles (UAVs) have rapidly emerged as an indispensable tool in precision agriculture due to their multiple advantages, including low cost, flexible deployment, high image resolution, real-time performance, and strong adaptability to complex terrains [1]. UAVs have demonstrated substantial value across various agricultural production stages, including field remote sensing monitoring, the early warning of pests and diseases, variable spraying and fertilization, crop growth modeling, and yield prediction, greatly advancing the intelligence level of agricultural management.

This Special Issue, entitled “Application of UAVs in Precision Agriculture—Part II”, includes nine representative research papers that systematically present recent advances in three key areas of UAV applications in agriculture: remote sensing detection technology focusing on efficient acquisition and the accurate interpretation of crop and environmental information; flight control technology aimed at improving UAV stability and task adaptability; and precision spraying technology oriented toward spatial optimization and the targeted application of agrochemical inputs. These studies comprehensively reflect the theoretical depth and technical breadth of current UAV applications in agriculture, from perception to decision-making and from platform optimization to operational implementation. Based on the contributions made to this Special Issue, this Editorial systematically reviews and analyzes the above three major directions, aiming to reveal development trends in UAV agricultural applications and provide a valuable technical framework and research perspective for researchers and practitioners in the field.

2. Overview of the Special Issue

2.1. Innovations in Remote Sensing Detection Technology

As a fundamental tool in precision agriculture, remote sensing technology plays a vital role in crop growth monitoring, pest and disease identification, and land cover extraction. Leveraging the high spatial and temporal resolution of UAVs, agricultural remote sensing has evolved from traditional macro-scale monitoring to high-precision field-scale identification and quantitative inversion [1]. In recent years, significant progress has been made in image preprocessing, feature extraction, object segmentation, and deep learning model development.

In the area of crop nutrient inversion, Bingquan Tian et al. proposed a method combining multispectral imagery with mixed-pixel decomposition, effectively improving the inversion accuracy of canopy chlorophyll-related indicators (SPAD values) in cotton. This study employed segmentation strategies including VIT, SVM, SMA, and MESMA to process remote sensing images and used modeling algorithms such as PLSR, RF, and SVR for multidimensional evaluation. The results confirmed that background removal and nonlinear modeling significantly enhance inversion performance, showing great field application potential [2].

To address the challenges of identifying agricultural targets in complex terrains, researchers have adopted deep learning methods to improve recognition accuracy. Xiaoyi Du et al. utilized a U-Net model to identify plastic film distribution in the Karst mountainous region. By optimizing model parameters and sample design, the model significantly improved area precision and object counting accuracy, showcasing the advantages of deep convolutional networks in identifying irregular coverage [3]. Similarly, Youyan Huang et al. combined geomorphic classification with the U-Net model to propose a multi-class segmentation approach for tobacco habitats. This method overcomes the limitations of traditional image processing in separating complex backgrounds, offering new data support and methodology for mountain agriculture management [4].

The influence of spatial resolution on image recognition accuracy has also gained attention. Xiandan Du et al. systematically analyzed the impact of flight altitude on image resolution and recognition performance, constructing a framework for Chinese cabbage plant identification using vegetation indices. This study found that an altitude of 30–50 m yields optimal recognition results, providing practical guidance for optimizing image acquisition strategies [5].

To meet the demands of large-scale, low-cost, and efficient agricultural monitoring, lightweight model development has become a research hotspot. Haoran Sun et al. proposed the P2P-CNF model, which maintains high counting accuracy for rice plants while significantly reducing model parameters and computational resources, validating the applicability of lightweight deep networks in agricultural scenarios [6].

2.2. Innovations in Flight Control Technology

The flight control system is central to ensuring efficient and safe UAV operation in agriculture, and its performance directly affects mission reliability and precision in complex environments. With agricultural operations expanding from flat fields to hilly and mountainous terrains, higher demands are placed on UAV maneuverability, adaptability, and interference resistance [7]. Consequently, various novel control strategies and structural designs have been proposed and validated in agricultural applications.

To address multi-scenario operation needs, Hao Qi et al. systematically studied the impact of free-tail layout parameters on the performance of tail-sitter vertical takeoff and landing (VTOL) UAVs. Compared with traditional configurations, this layout offers advantages in terrain adaptability and transition stability. Using the SST k-ω turbulence model for the numerical simulation of airflow fields, this study analyzed how different tail installation positions, lengths, deflection angles, and numbers affect flight performance. A multi-objective optimization method was used to determine the optimal layout, followed by flight validation. The results showed that the optimized layout significantly improved vertical stability, control responsiveness, and adaptability to complex terrain, providing a new structural design strategy for UAV applications in space-constrained environments like forests and mountains [8].

In another effort to enhance robustness and intelligence in flight control, Suiyuan Shen et al. proposed a composite control strategy integrating a fuzzy extended state observer (FESO) with sliding mode control (SMC). This approach treats system uncertainties and external disturbances as a total disturbance, which is estimated in real time by the FESO and compensated via the sliding mode control law. Simulations on an ALIGN E1 PLUS UAV platform showed that the FESO-SMC controller demonstrated stronger anti-disturbance capability and higher tracking accuracy than traditional controllers, particularly under sudden disturbances and parameter changes. This study offers important technical support for UAV stability in windy and interference-rich agricultural environments [9].

2.3. Advances in Precision Spraying Technology

Precision spraying is a critical part of crop protection, directly affecting pesticide utilization efficiency, control effectiveness, and environmental sustainability. UAVs are widely used in spraying due to their flexibility and high coverage efficiency, but factors such as wind disturbance, flight attitude changes, and crop structural variability impose challenges on droplet deposition [10]. Therefore, research into control algorithms and airflow modeling for optimizing spray effects has become essential.

Hang Xing et al. addressed efficiency fluctuations in rice spraying caused by wake vortices from multirotor UAVs by proposing a target control method based on a vortex–flight parameter model. By collecting flight data from E410 and Mavic 2 UAVs, this study established mappings between wake vortices and flight parameters and generated target parameter curves. Using Cube Orange flight controllers and Jetson AGX Xavier onboard systems, both PID and fuzzy controllers were designed and compared. The simulation results showed that the fuzzy controller exhibited better steady-state response and robustness in terms of height adjustment, flight speed, and vortex control, significantly reducing pesticide drift and deposition deviation, thus enhancing droplet targeting. This provides a technical foundation for intelligent UAV spraying systems and theoretical support for crop-specific flight strategy adjustments [11].

Meanwhile, Zongru Liu et al. focused on the relationship between UAV downwash airflow characteristics and spray deposition consistency. Using the k-ε turbulence model and dynamic mesh simulation, combined with wind field visualization experiments, this study analyzed the evolution of downwash geometries in DJI F450 UAVs. The results revealed a transition from “four-point” to “elliptical” airflow patterns under different flight conditions. Under low wind speeds, the airflow spread more rapidly, and deposition distribution was more uniform. Variations in nozzle position and local wind speed were also found to affect droplet deposition. This work validates the simulation model’s accuracy and provides experimental evidence for constructing generalized, predictive downwash models, which may be further refined using meteorological and crop structural data [12].

3. Conclusions

UAV technology, with its advantages in terms of efficiency, cost-effectiveness, flexible deployment, and high-resolution imaging, has become a key intelligent tool in precision agriculture. In remote sensing, the fusion of multispectral imagery with mixed-pixel decomposition significantly enhances the inversion of crop physiological parameters and object identification accuracy, especially under complex background conditions. The introduction of deep learning models such as U-Net enables the high-precision identification of plastic films and tobacco crops in heterogeneous and rugged terrains, reflecting a trend toward intelligent remote sensing analysis.

In the domain of flight control, efforts to optimize VTOL layout structures and incorporate fuzzy extended state observers with sliding mode control strategies have improved UAV maneuverability and robustness, ensuring reliable field operations.

In precision spraying, constructing vortex–flight parameter models and integrating fuzzy control algorithms has allowed for precise control over downwash airflow, while the in-depth analysis of airflow structures and droplet deposition has uncovered critical links between UAV spraying performance and crop protection outcomes.

In summary, UAV technologies have made substantial progress across the three key aspects of precision agriculture, showcasing the interdisciplinary synergy of innovations and steadily driving agricultural management from rough to refined and from experience-based to data-driven, providing strong support for high-quality agricultural development.