Drones, Volume 8, Issue 7 (July 2024) – 20 articles

Due to long-term exposure to the wild, insulators are prone to various defects that affect the safe operation of the power system. In recent years, the combination of drones and deep learning has provided a more intelligent solution for insulator automatic defect inspection. Positioning insulators is an important prerequisite step for defect detection, and the accuracy of insulator positioning greatly affects defect detection. However, traditional horizontal detectors lose directional information and it is difficult to accurately locate tilted insulators. Although oriented detectors can predict detection boxes with rotation angles to solve this problem, these models are complex and difficult to apply to edge devices with limited computing power. This greatly limits the practical application of deep learning methods in insulator detection. To address these issues, we proposed a lightweight insulator oriented detector. First, we designed a lightweight insulator feature pyramid network (LIFPN). It can fuse features more efficiently while reducing the number of parameters. Second, we designed a more lightweight insulator oriented detection head (LIHead). It has less computational complexity and can predict rotated detection boxes. Third, we deployed the detector on edge devices and further improved its inference speed through TensorRT. Finally, a series of experiments demonstrated that our method could reduce the computational complexity of the detector by approximately 49 G and the number of parameters by approximately 30 M while ensuring almost no decrease in the detection accuracy. It can be easily deployed to edge devices and achieve a detection speed of 41.89 frames per second (FPS). Full article

Invasive knotweeds are rhizomatous and herbaceous perennial plants that pose significant ecological threats due to their aggressive growth and ability to outcompete native plants. Although detecting and identifying knotweeds is crucial for effective management, current ground-based survey methods are labor-intensive and limited to cover large and hard-to-access areas. This study was conducted to determine the optimum flight height of drones for aerial detection of knotweeds at different phenological stages and to develop automated detection of knotweeds on aerial images using the state-of-the-art Swin Transformer. The results of this study found that, at the vegetative stage, Japanese knotweed and giant knotweed were detectable at ≤35 m and ≤25 m, respectively, above the canopy using an RGB sensor. The flowers of the knotweeds were detectable at ≤20 m. Thermal and multispectral sensors were not able to detect any knotweed species. Swin Transformer achieved higher precision, recall, and accuracy in knotweed detection on aerial images acquired with drones and RGB sensors than conventional convolutional neural networks (CNNs). This study demonstrated the use of drones, sensors, and deep learning in revolutionizing invasive knotweed detection. Full article

Marine aerosol particles can act as cloud condensation nuclei and influence the atmospheric boundary layer by scattering solar radiation. The interaction of ocean waves and coral reefs may affect the distribution and size of marine aerosol particles. Measuring this effect has proven challenging. Here, we tested the hypothesis that the distribution and size of marine aerosol particles would vary over three distinct zones (i.e., coral lagoon, surf break, and open water) near One Tree Island in the Great Barrier Reef, which is approximately 85 km off the east coast of Australia. We used a modified DJI Agras T30 drone fitted with a miniaturised scanning electrical mobility sizer and advanced mixing condensation particle counter to collect data on aerosol size distribution between 30 and 300 nm at 20 m above the water surface. We conducted 30 flights over ten days during the Austral summer/autumn of 2023. The fitted bimodal lognormal curves indicate that the number concentrations for aerosols below 85 nm diameter are more than 16% higher over the lagoon than over open water. The average mean mode diameters remained constant across the different zones, indicating no significant influence of breaking waves on the detected aerosol size modes. The most influential explanatory variable for aerosol size distribution was the difference between air temperature and the underlying sea surface, explaining around 40% of the variability. Salinity also exhibited a significant influence, explaining around 12% of the measured variability in the number concentration of aerosols throughout the campaign. A calculated wind stress magnitude did not reveal significant variation in the measured marine aerosol concentrations. Overall, our drone-based aerosol measurements near the water surface effectively characterise the dynamics of background marine aerosols around One Tree Island Reef, illustrating the value of drone-based systems for providing size-dependent aerosol information in difficult-to-access and environmentally sensitive areas. Full article

The aim of this paper is to identify the possibilities of the implementation of the Innovative Management Framework for Intelligent Horticulture (IMFIH) by farmers with the aim of deepening the dynamics of innovation and technologic transfer processes related to the integration of the aerial work capability offered by mini UAV systems in precision horticulture. Starting from an aerial system for intelligent monitoring and smart horticulture applications, the research methodology is designed to understand the specific processes of this transfer of innovation in a field characterized by evolutionary dynamics and in the context of a lack of data. Thus, it is considered to be a mix of both quantitative and qualitative methods used in order to identify the needs and opinions of farmers regarding the possible use of the capabilities of mini UAV systems and especially how to access this capability. The obtained results showed the profile of the farmers interested in mini UAV systems for monitoring field crops and also the relevant factors for initiating/accessing them: specialized education, entrepreneurial education, area owned, ability to create partnerships, intention to access/develop mini UAV systems, and the existence of an integrated framework for analyzing the opportunities and restrictions of implementing mini UAV systems in precision horticulture applications. The integration of the Hype Cycle Paradigm (HCP) into the proposal of IMFIH led to the creation of the IMFIH-HCP as an innovative framework capable of stimulating the dissemination and transfer of knowledge and technology in the case of future horticultural applications of precision in an emerging market. Full article

Knowledge on the spatiotemporal patterns of surface energy balance parameters is crucial for understanding climate system processes. To this end, the assimilation of Earth Observation data with land biosphere models has shown promising results, but they are still hampered by several limitations related to the spatiotemporal resolution of EO sensors and cloud contamination. With the recent developments on Unmanned Aerial Vehicles (UAVs), there is a great opportunity to overcome these challenges and gain knowledge of surface energy balance parameters at unprecedented resolutions. The present study examines, for the first time, the ability of an inversion-modeling scheme, the so-called “analytical triangle” method, to retrieve estimates of surface energy fluxes and soil surface moisture (SSM) at high spatial resolution using UAV data. A further aim of our study was to examine the representativeness of the SSM estimates for the SM measurements taken at different depths. The selected experimental site is an agricultural site of citrus trees located near the city of Palermo on 30 July 2019. The results of comparisons showed that the sensible and latent heat fluxes from UAV were consistent with those measured from the ground, with absolute differences in comparison to ground measurements being 5.00 Wm⁻² for the latent heat (LE) flux and 65.02 Wm⁻² for H flux, whereas for the daytime fluxes H/Rn and LE/Rn were 0.161 and 0.012, respectively. When comparing analytical triangle SSM estimates with SM measurements made at different depths, it was found that there was a gradual increase in underestimation with increasing measurement depth. All in all, this study’s results provide a credible demonstration of the significant potential of the technique investigated herein as a cost-effective and rapid solution for estimating key parameters characterizing land surface processes. As those parameters are required by a wide range of disciplines and applications, utilization of the investigated technique in research and practical applications is expected to be seen in the future. Full article

Unmanned aerial vehicles (UAV) have been used in many fields nowadays. In long-term applications, batteries need to be constantly changed by someone due to short battery life. This problem is eliminated with wireless power transfer (WPT). A reliable, effective, and autonomous solution is offered using wireless charging. The most suitable wireless charging technique for UAVs is inductive power transfer (IPT). In this paper, a novel foldable coil and charge station design is proposed for the wireless charging of UAVs. IPT is provided by receiver and transmitter coils placed on the drone legs and the charging station, respectively. Receiver coils are placed on both legs of the UAV in a light and balanced manner to avoid creating imbalance and weight on the UAV. Receiver coils are designed as vertical rectangular planar spirals. A transmitter coil consists of three rectangular planar spiral coils with two movable edge windings and a fixed middle winding. The transmitter’s folding windings provide both alignments for the UAV during landing and increase the magnetic coupling. A folding wireless charge system of the UAV is designed for 100 W output power at a 138.1 kHz frequency. The misalignment tolerance of the proposed design in the vertical axis is examined. The design’s magnetic flux density distribution is analysed. As an experimental result of the study, 97.66% efficiency was reached in the aligned condition. Also, over 85.48% efficiency was achieved for up to 10 cm of vertical alignment misalignment. Full article

This paper introduces a comprehensive framework for generating obstacle-free flight paths for unmanned aerial vehicles (UAVs) in intricate 3D environments. The system leverages the Rapidly Exploring Random Tree (RRT) algorithm to design trajectories that effectively avoid collisions with structures of diverse shapes and sizes. Discussion revolves around the challenges encountered during development and the successful achievement of generating collision-free routes. While the system represents an initial iteration, it serves as a foundation for future projects aiming to refine and expand upon its capabilities. Future work includes simulation testing and integration into UAV missions for image acquisition and structure scanning. Additionally, considerations for swarm deployment and 3D reconstruction using various sensor combinations are outlined. This research contributes to the advancement of autonomous UAV navigation in real-world scenarios. Full article

Crop above-ground biomass (AGB) estimation is a critical practice in precision agriculture (PA) and is vital for monitoring crop health and predicting yields. Accurate AGB estimation allows farmers to take timely actions to maximize yields within a given growth season. The objective of this study is to use unmanned aerial vehicle (UAV) multispectral imagery, along with derived vegetation indices (VI), plant height, leaf area index (LAI), and plant nutrient content ratios, to predict the dry AGB (g/m²) of a winter wheat field in southwestern Ontario, Canada. This study assessed the effectiveness of Random Forest (RF) and Support Vector Regression (SVR) models in predicting dry ABG from 42 variables. The RF models consistently outperformed the SVR models, with the top-performing RF model utilizing 20 selected variables based on their contribution to increasing node purity in the decision trees. This model achieved an R² of 0.81 and a root mean square error (RMSE) of 149.95 g/m². Notably, the variables in the top-performing model included a combination of MicaSense bands, VIs, nutrient content levels, nutrient content ratios, and plant height. This model significantly outperformed all other RF and SVR models in this study that relied solely on UAV multispectral data or plant leaf nutrient content. The insights gained from this model can enhance the estimation and management of wheat AGB, leading to more effective crop yield predictions and management. Full article

This paper proposes a systematic approach for identifying the translational dynamics of a novel two-layer octocopter. Initially, we derive the non-linear theoretical dynamic model of the conventional octocopter using the Newton–Euler formulation, aimed at obtaining a simplified model suitable for tuning PID gains necessary for controller implementation. Following this, a controller is designed and tested in the Matlab/Simulink environment to ensure stable flight performance of the octocopter. Subsequently, the novel octocopter prototype is developed, fabricated, and assembled, followed by a series of outdoor flight tests conducted under various environmental conditions to collect data representing the flight characteristics of the two-layer vehicle in different scenarios. Based on the data recorded during flights, we identify the transfer functions of the translational dynamics of the modified vehicle using the prediction error method (PEM). The empirical model is then validated through different flight tests. The results presented in this study exhibit a high level of agreement and demonstrate the efficacy of the proposed approach to predict the octocopter’s position based only on motor inputs and initial states of the system. Despite the inherent non-linearity, significant aerodynamic interactions, and strongly coupled nature of the system, our findings highlight the robustness and reliability of the proposed approach, which can be used to identify the model of any type of multi-rotor or fixed-wing UAV, specifically when you have a challenging design. Full article

Time synchronization is a critical prerequisite for unmanned aerial vehicle ad hoc networks (UANETs) to facilitate navigation and positioning, formation control, and data fusion. However, given the dynamic changes in UANETs, improving the convergence speeds of distributed consensus time synchronization algorithms with only local information poses a major challenge. To address this challenge, this study first establishes a convex model on the basis of graph theory and relevant theories of random matrices to approximate the original problem. Subsequently, three acceleration schemes for consensus algorithms are derived by minimizing the Frobenius norm of the iteration matrix. Additionally, this study provides a new upper bound for constant communication weights and discusses the limitations of existing metrics used to measure the convergence speeds of consensus algorithms. Finally, the proposed schemes are compared with existing ones through simulation. Our results indicate that the three proposed schemes can achieve faster convergence while maintaining high-precision synchronization in scenarios with static or known topological structures of networks. In scenarios where the topological structure of a UANET is time-varying and unknown, the scheme proposed in this paper achieves the fastest convergence speed. Full article

Accurate forecasting of crop yields holds paramount importance in guiding decision-making processes related to breeding efforts. Despite significant advancements in crop yield forecasting, existing methods often struggle with integrating diverse sensor data and achieving high prediction accuracy under varying environmental conditions. This study focused on the application of multi-sensor data fusion and machine learning algorithms based on unmanned aerial vehicles (UAVs) in wheat yield prediction. Five machine learning (ML) algorithms, namely random forest (RF), partial least squares (PLS), ridge regression (RR), k-nearest neighbor (KNN) and extreme gradient boosting decision tree (XGboost), were utilized for multi-sensor data fusion, together with three ensemble methods including the second-level ensemble methods (stacking and feature-weighted) and the third-level ensemble method (simple average), for wheat yield prediction. The 270 wheat hybrids were used as planting materials under full and limited irrigation treatments. A cost-effective multi-sensor UAV platform, equipped with red–green–blue (RGB), multispectral (MS), and thermal infrared (TIR) sensors, was utilized to gather remote sensing data. The results revealed that the XGboost algorithm exhibited outstanding performance in multi-sensor data fusion, with the RGB + MS + Texture + TIR combination demonstrating the highest fusion performance (R² = 0.660, RMSE = 0.754). Compared with the single ML model, the employment of three ensemble methods significantly enhanced the accuracy of wheat yield prediction. Notably, the third-layer simple average ensemble method demonstrated superior performance (R² = 0.733, RMSE = 0.668 t ha⁻¹). It significantly outperformed both the second-layer ensemble methods of stacking (R² = 0.668, RMSE = 0.673 t ha⁻¹) and feature-weighted (R² = 0.667, RMSE = 0.674 t ha⁻¹), thereby exhibiting superior predictive capabilities. This finding highlighted the third-layer ensemble method’s ability to enhance predictive capabilities and refined the accuracy of wheat yield prediction through simple average ensemble learning, offering a novel perspective for crop yield prediction and breeding selection. Full article

Aerial robots (drones) offer critical advantages in missions where human participation is impeded due to hazardous conditions. Among these, search and rescue missions in disaster-stricken areas are particularly challenging due to the dynamic and unpredictable nature of the environment, often compounded by the lack of reliable environmental models and limited ground system communication. In such scenarios, autonomous aerial robots’ operation becomes essential. This paper introduces a novel hierarchical reinforcement learning-based algorithm to address the critical limitation of the aerial robot’s battery life. Central to our approach is the integration of a long short-term memory (LSTM) model, designed for precise battery consumption prediction. This model is incorporated into our HRL framework, empowering a high-level controller to set feasible and energy-efficient goals for a low-level controller. By optimizing battery usage, our algorithm enhances the aerial robot’s ability to deliver rescue packs to multiple survivors without the frequent need for recharging. Furthermore, we augment our HRL approach with hindsight experience replay at the low level to improve its sample efficiency. Full article

The spread of natural fires is a complex issue, as its mathematical modeling needs to consider many parameters. Therefore, the results of such modeling always need to be validated by comparison with experimental measurements under real-world conditions. Remote sensing with the support of satellite or aerial sensors has long been used for this purpose. In this article, we focused on data collection with an unmanned aerial vehicle (UAV), which was used both for creating a digital surface model and for dynamic monitoring of the spread of controlled grassland fires in the visible spectrum. We subsequently tested the impact of various processing settings on the accuracy of the digital elevation model (DEM) and orthophotos, which are commonly used as a basis for analyzing fire spread. For the DEM generated from images taken during the final flight after the fire, deviations did not exceed 0.1 m compared to the reference model from LiDAR. Scale errors in the model with only approximal WGS84 exterior orientation parameters did not exceed a relative accuracy of 1:500, and possible deformations of the DEM up to 0.5 m in height had a minimal impact on determining the rate of fire spread, even with oblique images taken at an angle of 45°. The results of the experiments highlight the advantages of using low-cost SfM photogrammetry and provide an overview of potential issues encountered in measuring and performing photogrammetric processing of fire spread. Full article

With the frequent occurrence of drone-related criminal cases, drone forensics has become a hot spot of concern. During drone-related criminal investigations, the implicated drones are often forcibly brought down, which poses significant challenges in conducting forensic analysis. In order to restore the truth of criminal cases, it is necessary to extract data not only from the external TF card but also from internal chip memory in drone forensics. To address this issue, a drone data parser (DRDP) is proposed to extract internal and external data from criminal-implicated drones. In this paper, we present comprehensive forensics on the DJI Mavic 2 Pro, analyzing the main file structure and encryption model. According to its file structures, three case studies are conducted on various file types (DAT files, TXT files, and default files) to verify the effectiveness and applicability of the designed procedure. The results show that the encrypted data of the implicated drone, such as GPS information, flight time, flight altitude, flight distance, three velocity components (x, y, z) and other information can be extracted and decrypted correctly, which provides evidence for the identification of the case facts. Full article

This article deals with quadcopter modeling using a system for the measurement of unmanned aerial vehicle (UAV) parameters. UAVs are often equipped with various measurement devices and equipment for measurement, which significantly affects their weight. The currently available technical solutions and inventions do not allow corrections to be made to the on-board control electronics settings without the need to perform a test flight, or without the need to create complex and time-consuming mathematical models of the unmanned aerial vehicle; therefore, it is desirable to create a new method for modeling the characteristics of an UAV based on static laboratory measurements. The goal of this paper is to create a dynamic model of a quadcopter that will be adapted to a system for measuring UAV parameters, specifically the thrust of individual motors, which will be the next step to creating a new method for modeling UAV characteristics. This method can be used in the future for tuning flight control algorithms, based on static laboratory measurements. Full article

Unmanned aerial vehicles (UAVs)-based Person Re-Identification (ReID) is a novel field. Person ReID is the task of identifying individuals across different frames or views, often in surveillance or security contexts. At the same time, UAVs enhance person ReID through their mobility, real-time monitoring, and ability to access challenging areas despite privacy, legal, and technical challenges.To facilitate the advancement and adaptation of existing person ReID approach to the UAV scenarios, this paper introduces a baseline along with two datasets, i.e., LSMS and LSMS-UAV. Both datasets have the following key features: (1) LSMS: Raw videos captured by a network of 29 cameras deployed across complex outdoor environments. LSMS-UAV: captured by 1 UAV. (2) LSMS: Videos span both winter and spring seasons, encompassing diverse weather conditions and various lighting conditions throughout different times of the day. (3) LSMS: Including the largest number of annotated identities, comprising 7730 identities and 286,695 bounding boxes. LSMS-UAV: comprising 500 identities and 2000 bounding boxes. Comprehensive experiments demonstrate LSMS’s excellent capability in addressing the domain gap issue when facing complex and unknown environments. The LSMS-UAV dataset verifies that UAV data has strong transferability to traditional camera-based data. Full article

This work presents a minimum separation calculation for the integrated operation of manned and unmanned aerial vehicles in an uncertain airspace environment. Different from traditional path-planning-based research, this study investigated the minimum safe separation distance from a novel perspective of reachability analysis. The proposed computational method made use of the Hamilton–Jacobi partial differential equation (HJPDE) to obtain the backward reachable tube. Firstly, this work modeled the integrated operation in the UAS traffic management scenario, particularly focusing on the uncertainties. Then, a probabilistic reachability tube computation method was derived. Next, this work calculated the safe separation distances based on reachability analysis for three scenarios: a deterministic environment, an environment with relative position uncertainty, and an environment with relative heading angle uncertainty. By calculating the reachable tubes for a given response time, the worst-case minimum safe distances from the UAV’s perspective were determined, and the quantitative patterns were summarized. The results in this work indicate that, with an increase in the risk level and under the premise of a 1 second response time, the minimum safe separation increases from 26.7 meters to 30.0 meters. Finally, the paper discusses the results, explaining their rationality from both mathematical and physical perspectives. Full article

The precise regulation of the hinge moment and pitch angle driven by the pitch-regulated mechanism is crucial for modulating thrust requirements and ensuring stable attitude control in Martian coaxial rotorcraft. Nonetheless, the aerodynamic hinge moment in rotorcraft presents time-dependent dynamic properties, posing significant challenges for accurate measurement and assessment for such characteristics. In this study, we delve into the detailed aerodynamic hinge moment characteristics associated with the pitch-regulated mechanism of Mars rotorcraft under a spectrum of control strategies. A robust computational fluid dynamics model was developed to simulate the rotor’s aerodynamic loads, accompanied by a quantitative hinge moment characterization that takes into account the effects of varying rotor speeds and pitch angles. Our investigation yielded a thorough understanding of the interplay between aerodynamic load behavior and rotor surface pressure distributions, leading to the creation of an empirical mapping model for hinge moments. To validate our findings, we engineered a specialized test apparatus capable of measuring the hinge moments of the pitch-regulated mechanism, facilitating empirical assessments under replicated atmospheric conditions of both Earth and Mars. The result indicates aerodynamic hinge moments depend nonlinearly on rotational speed, peaking at a 0° pitch angle and showing minimal sensitivity to pitch under 0°. Above 0°, hinge moments decrease, reaching a minimum at 15° before rising again. Simulation and experimental comparisons demonstrate that under Earth conditions, the aerodynamic performance and hinge moment errors are within 8.54% and 24.90%, respectively. For Mars conditions, errors remain below 11.62%, proving the CFD model’s reliability. This supports its application in the design and optimization of Mars rotorcraft systems, enhancing their flight control through the accurate prediction of aerodynamic hinge moments across various pitch angles and speeds. Full article

The task of multiple-tiny-object detection from diverse perspectives in unmanned aerial vehicles (UAVs) using onboard edge devices is a significant and complex challenge within computer vision. In order to address this challenge, we propose a lightweight and efficient tiny-object-detection algorithm named LE-YOLO, based on the YOLOv8n architecture. To improve the detection performance and optimize the model efficiency, we present the LHGNet backbone, a more extensive feature extraction network, integrating depth-wise separable convolution and channel shuffle modules. This integration facilitates a thorough exploration of the inherent features within the network at deeper layers, promoting the fusion of local detail information and channel characteristics. Furthermore, we introduce the LGS bottleneck and LGSCSP fusion module incorporated into the neck, aiming to decrease the computational complexity while preserving the detector’s accuracy. Additionally, we enhance the detection accuracy by modifying its structure and the size of the feature maps. These improvements significantly enhance the model’s capability to capture tiny objects. The proposed LE-YOLO detector is examined in ablation and comparative experiments on the VisDrone2019 dataset. In contrast to YOLOv8n, the proposed LE-YOLO model achieved a 30.0% reduction in the parameter count, accompanied by a 15.9% increase in the mAP(0.5). These comprehensive experiments indicate that our approach can significantly enhance the detection accuracy and optimize the model efficiency through the organic combination of our suggested enhancements. Full article

With the development and strengthening of interception measures, the traditional penetration methods of high-speed unmanned aerial vehicles (UAVs) are no longer able to meet the penetration requirements in diversified and complex combat scenarios. Due to the advancement of Artificial Intelligence technology in recent years, intelligent penetration methods have gradually become promising solutions. In this paper, a penetration strategy for high-speed UAVs based on improved Deep Reinforcement Learning (DRL) is proposed, in which Long Short-Term Memory (LSTM) networks are incorporated into a classical Soft Actor–Critic (SAC) algorithm. A three-dimensional (3D) planar engagement scenario of a high-speed UAV facing two interceptors with strong maneuverability is constructed. According to the proposed LSTM-SAC approach, the reward function is designed based on the criteria for successful penetration, taking into account energy and flight range constraints. Then, an intelligent penetration strategy is obtained by extensive training, which utilizes the motion states of both sides to make decisions and generate the penetration overload commands for the high-speed UAV. The simulation results show that compared with the classical SAC algorithm, the proposed algorithm has a training efficiency improvement of 75.56% training episode reduction. Meanwhile, the LSTM-SAC approach achieves a successful penetration rate of more than 90% in hypothetical complex scenarios, with a 40% average increase compared with the conventional programmed penetration methods. Full article