UAV Technology Research Articles

Research on Recognition Method of Traffic Sign Images Based on UAV Technology

Research on Recognition Method of Traffic Sign Images Based on UAV Technology

Viabilidade do uso de drone aplicado no mapeamento da pesca artesanal comercial: um estudo de caso na Baía de Sepetiba, Rio de Janeiro

ABSTRACT: This article presents an experimental study to verify the feasibility of using the emerging UAV technology in the mapping of fishing grounds exploited by artisanal fishing in Sepetiba Bay, in the south of the state of Rio de Janeiro-Brazil, aiming to better understand the functionality and information generated by the equipment, and whether the data can help improve local fisheries management and also provide information for the implementation of simplified traceability for regional fish. The study area included four locations, Ilha da Madeira, Vila Geny, Coroa Grande and Ponte Preta, all located in Sepetiba Bay, located in the municipality of Itaguaí, in the southern region of the state of Rio de Janeiro. The aerial mapping was carried out in September, October and November 2017, between 10 am and 12 pm (GMT-3), employing a Phantom 3 Professional UAV (DJI, Shenzhen, China), Anac Brasil registration number PP-011092014. The onboard camera a DJI 4K Edition Sony Exmor R Model IMX117: 7.81mm equipped with a CMOS sensor dimension of 6.2 mm, 4000 x 3000 12 megapixel resolution, with an f/2.8 lens and field of view (FOV) of 94, having a focal length of 14mm. Images were collected perpendicularly to the main plan. Flight height was set at 80 m, flight speed was determined based on the overlap of 80%, dimension of area to be mapped, and battery life. Hence, the operator utilized a average flight speed of 25 km/h. ground sample distance (GSD) was 1.20 cm/px. and an overlap of 80% between images was established to avoid possible failures and orthomosaic information gaps. For each fishery mapped, standard software processing was carried out, which consists of: aligning the photos, creating a point cloud, creating a digital elevation model, and finally the orthomosaic itself. Once processed, the orthomosaics were exported from Agis Soft Photo Scan Professional to Google Earth (Google Inc.), where it is possible to make a more comprehensive assessment of fisheries in a much more practical and faster way. A total of four flights were carried out, one for each study area, averaging 10 minutes each. As a result, the technology proved to be viable to apply, since it made it possible to identify and map the areas defined for this research, generating data such as distance and, characteristics of the place, which could be used to track the fish caught and landed, in addition to enabling the monitoring of vessels and fishing, and thus improve the management of local fisheries.

LDHD‐Net: A Lightweight Network With Double Branch Head for Feature Enhancement of UAV Targets in Complex Scenes

The development of small UAV technology has led to the emergence of new challenges in UAV countermeasures. The timely detection of UAVs can effectively prevent potential infringements on airspace and privacy. Currently, methods based on deep learning demonstrate excellent performance in target detection. However, in complex scenes, there is a tendency for false alarms (FAs) and misdetections to occur at a higher rate. To solve these problems, we propose a lightweight infrared small target detection algorithm LDHD‐Net. First, we design a novel Ghost‐Shuffle module in the backbone network to enhance the network feature extraction capability. Meanwhile, we remove redundant layers from the network to make the backbone network more lightweight. Second, we design a hierarchical attention enhancement module in the neck network to improve the saliency of UAV targets and reduce background noise interference. In addition, we design a novel small target detection structure and prediction heads in the shallow layers of the network to improve small target detection accuracy. Finally, we design a novel attention dual‐branch head to reduce interference between different tasks and improve the real‐time performance of algorithm detection. The experimental results show that compared with the original model, inference time remains essentially the same, LDHD‐Net parameters are only 3.9 M and AP improves by 12.6%. Compared to SOTA methods, LDHD‐Net shows better performance on SIDD and Anti‐UAV410 datasets. The algorithm effectively improves the accuracy and real‐time detection of UAVs in complex scenes.

Enhancing unmanned aerial vehicle capabilities: integrating YOLO algorithms for diverse industrial applications

The integration of UAVs with advanced deep learning algorithms, particularly the You Only Look Once models, has opened new horizons in various industries. This paper explores the transformative impact of YOLO-based systems across diverse sectors, including agriculture, forest fire detection, ecology, marine science, target detection, and UAV navigation. We delve into the specific applications of different YOLO models, ranging from YOLOv3 to the lightweight YOLOv8, highlighting their unique contributions to enhancing UAV functionalities. In agriculture, UAVs equipped with YOLO algorithms have revolutionized disease detection, crop monitoring, and weed management, contributing to sustainable farming practices. The application in forest fire management showcases the capability of these systems in real-time fire localization and analysis. In ecological and marine sciences, the use of YOLO models has significantly improved wildlife monitoring, environmental surveillance, and resource management. Target detection studies reveal the efficacy of YOLO models in processing complex UAV imagery for accurate and efficient object recognition. Moreover, advancements in UAV navigation, through YOLO-based visual landing recognition and operation in challenging environments, underscore the versatility and efficiency of these integrated systems. This comprehensive analysis demonstrates the profound impact of YOLO-based UAV technologies in various fields, underscoring their potential for future innovations and applications.

Smart Farming: Integrating IoT and UAV Technologies for Precision Agriculture through the Lens of Technology Acceptance and the UTAUT2 Model

Smart Farming: Integrating IoT and UAV Technologies for Precision Agriculture through the Lens of Technology Acceptance and the UTAUT2 Model

A Systematic Review of the UAV Technology Usage in ASEAN

A Systematic Review of the UAV Technology Usage in ASEAN

Evaluation of rockfall hazard based on UAV technology and 3D Rockfall Simulations

Hacıabdullah village, where rockfall events occur frequently, is an important mountainous region located within the boundaries of Niğde province, which is one of the top provinces in Türkiye in terms of rockfall events and is located in the Central Anatolia Region. Despite the frequent occurrence of many rockfall events of varying sizes, no precautions are taken as a result of these events, leading to economic damages. Therefore, Hacıabdullah village, which poses a high danger and risk in terms of slope instability and rockfall potential, has been selected as the study area. The study aims to evaluate rockfall events occurring on steep slopes in Haciabullah village, based on the potential rock blocks that could fall, by modeling them according to their geometries. Within this scope, the potential 17 rock blocks that could fall have been identified in the field and modelled using RAMMS 3D rockfall software according to their geometric characteristics. In addition to the geometric characteristics of the rock blocks, their potential falling areas, velocities, jump heights, and kinetic energies will also be determined, the basis of the rockfall hazard analysis. For the 3D modeling of the study, UAV-DSM (3cm resolution), which represents the high-resolution surface of the topography, has been used as the main dataset. As a result of 3D rockfall modeling, the maximum kinetic energy, maximum velocity, and maximum jump height of the falling blocks reached 3476 kJ, 23.1 m/s, and 14.57 m, respectively. The result of the kinematic analysis showed a higher probability of toppling type in the whole study area. Rocks that do not move very far from the source area are; in other words and may significantly damage the roads. However, rolling blocks, in other words, blocks that can travel long distances from the source area, have the potential to cause great damage to the settlement areas, roads, and trees. According to the hazard map, blocks involve high and moderate levels of risk for settlement units.

Real-Time Object Detection and Tracking for Unmanned Aerial Vehicles Based on Convolutional Neural Networks

This paper presents a system applied to unmanned aerial vehicles based on Robot Operating Systems (ROSs). The study addresses the challenges of efficient object detection and real-time target tracking for unmanned aerial vehicles. The system utilizes a pruned YOLOv4 architecture for fast object detection and the SiamMask model for continuous target tracking. A Proportional Integral Derivative (PID) module adjusts the flight attitude, enabling stable target tracking automatically in indoor and outdoor environments. The contributions of this work include exploring the feasibility of pruning existing models systematically to construct a real-time detection and tracking system for drone control with very limited computational resources. Experiments validate the system’s feasibility, demonstrating efficient object detection, accurate target tracking, and effective attitude control. This ROS-based system contributes to advancing UAV technology in real-world environments.

The Classification Characteristics and Dynamic Changes of Desert Vegetation Based on Unmanned Aerial Vehicle Remote Sensing

The common ground sample method of vegetation survey requires a lot of manpower and time. Meanwhile, many vegetation information may not be revealed because the scale of standard sample is usually only 100 m2. For this reason, this study developed a set of desert vegetation sample survey method based on UAV technology, and further developed a set of vegetation spatial distribution pattern analysis method based on it, and carried out an empirical study with typical shrub communities in Inner Mongolia as an example. It is expected to provide scientific basis for wind and sand control, vegetation restoration and ecological construction in the ecologically fragile area at the edge of Inner Mongolia.

Morphometric Parameters of the Water Bodies of Bagry Wielkie and Bagry Małe in the Biodiversity Context

The article presents the results of nature-based and geodetic query of two neighbouring endorheic borrow pits, i.e., Bagry Wielkie and Staw Płaszowski (Bagry Małe), situated almost in the centre of the Krakow city. The movements of water tables in 2016–2023 were observed (from the moment of ecological disaster in the Staw Płaszowski, in February 2016) and bathymetric maps were presented. Based on geodetic measurements carried out in 2016–2022 and direct observations of nature, documenting biodiversity of both reservoirs in 2019–2022, the authors reviewed the state of biosphere in this location. Surveying, observations and interviews allowed parameterization of both water bodies (morphometric data and bathymetric maps) and creating up-to-date database (location on maps, qualitative and quantitative elements of flora and fauna). The material makes base for further studies with the application of more advanced photogrammetric technologies and UAV technologies, which can make data contained in this publication more accurate.

Study on The Applicability of Drones in The Process of Forming Towers of 500KV Transmission and Substation Lines in Highland Areas

With the continuous development of science and technology in China, China's aviation field has also been expanding, the rapid development of aviation technology, China has gradually achieved leapfrog development in UAV equipment, and gradually tends to civilization, practicality and commercialization. At present, more and more UAVs are not only opening up in the domestic market, but also have flowed to overseas markets, and the modern advanced technology represented by UAVs has gradually become the current trend development trend. In the field of electric power engineering construction, the use of drone technology is short and wide, and there are drones in both high-altitude areas and plain areas. In terms of the engineering field, the UAV technology involves the construction, monitoring and management of power grids, fault and equipment maintenance of power grids and transmission line inspection, etc.

A Model Design in UAV Technology: A Multi-controller-based Wi-Fi Imaging Drone

A Model Design in UAV Technology: A Multi-controller-based Wi-Fi Imaging Drone

A Conceptual Framework for Economic Analysis of Different Law Enforcement Drones

The widespread use of drones in various fields has initiated a discussion on their cost-effectiveness and economic impact. This article analyzes in detail a methodological evaluation framework for the levelized cost of drone services for law enforcement purposes. Based on the data availability, we compared two vehicles: Phantom 4 Pro and Thunder-B. Moreover, we calculated their levelized costs per surveillance time and trip distance. Our approach helps users calculate the real costs of their vehicles’ services and produce equations for rapid estimations. We observed economies of scale for time and distance and showed differentiations per aircraft capacity. Furthermore, using the produced equations, we formulated a case study and compared the costs in a 4 km area constantly monitored by the two types of drones to support the best vehicle selection. We found that the Phantom 4 Pro costs less than the Thunder-B drone, for example. Thus, we demonstrate how, by applying this methodology beforehand, decision makers can select the most appropriate vehicle for their needs based on cost. Cost research estimations will improve UAV use and will help policymakers include UAV technology in crime prevention programs, especially when more data are available.

A multi-strategy particle swarm algorithm with exponential noise and fitness-distance balance method for low-altitude penetration in secure space

UAV technology is at the forefront of current research and plays an important role in agriculture, public safety and the military. Low altitude trajectory is a technology for UAVs to fly safely in situations full of various threats, so the safety and stability of the flight path is an important challenge. Current swarm intelligence (SI) algorithms are widely used in path planning problems due to their high robustness and global search capability, but most of them do not take into account the universality of such problems and flight control in safe space, so it is particularly important to improve the ability of SI to plan paths in multiple environments. To better improve the competitiveness of SI in such problems, this paper proposes a multi-strategy particle swarm algorithm based on exponential noise to solve low altitude penetration in secure space, and the algorithm is abbreviated as MEPSO. A selective opposition (SO) strategy is introduced in the traditional particle swarm algorithm to dynamically update the position of individuals and improve the flexibility of the algorithm; then uses exponential noise and fitness-distance balance (FDB) to expand the search space and prevent the algorithm from falling into a local optimum, further enhance the optimization capability of the algorithm. Comparing MEPSO with 4 basic algorithms and 7 variants of other algorithms, the results show that the minimum and mean costs of MEPSO in simple environments are 72.5850 and 72.7472, respectively. It also shows that the minimum and mean cost in complex environments are 73.2984 and 73.3312, respectively. MEPSO has better global optimization capability than other algorithms. The safety and stability of the paths in both environments are high, which verifies that the path planning capability of MEPSO is reasonable and effective in the low altitude penetration in secure space.

Prediction Parameters for Mining Subsidence Based on Interferometric Synthetic Aperture Radar and Unmanned Aerial Vehicle Collaborative Monitoring

Coal mining induces surface subsidence, making rapid and precise monitoring of this subsidence a key area of current research. To address the limitations of D-InSAR technology in capturing large-gradient deformations in the central subsidence basin and the challenges facing UAVs in accurately monitoring small deformations at the basin’s edge, we propose a method for inverting the expected parameters of surface subsidence by synergistically integrating InSAR and UAV monitoring. We determined the cumulative subsidence of monitoring points along the dip and strike observation line of the Banji 110,801 working face between 10 April 2021 and 28 June 2022, employing D-InSAR and UAV techniques. By leveraging the complementary strengths of both monitoring techniques, we fused the two types of monitoring data and verified the error of the fusion data to be within 10 cm through leveling data verification. Simulation experiments utilizing the probability integration method and the Broyden–Fletcher–Goldfarb–Shanno (BFGS) optimization algorithm confirmed that the 10 cm data source error remains within the required limits for probability integration parameter inversion. Finally, the BFGS algorithm was employed to invert the parameters of the probability integration method based on the fusion data results. Subsequently, these inversion parameters were used to predict the subsidence at the monitoring point and were compared with the level measured data. The results demonstrate that the use of collaborative InSAR and UAV monitoring technology for inverting the expected parameters of surface subsidence in the mining area yields superior results, aligning with the actual patterns of ground surface movement and deformation. This study addresses the global need for unmanned monitoring of mining-related subsidence. It employs InSAR and UAV technologies in a synergistic approach to monitor surface subsidence in mining regions. This approach harnesses the strengths of multiple data sources and presents a novel concept for the unmanned monitoring of surface subsidence in mining areas, contributing to environmental protection efforts.

Research on the Application Status of UAV in Agriculture

UAVs are widely used in agriculture. Through high-resolution image and data collection, they help farmers carry out crop monitoring and evaluation, precision agriculture, plant protection and crop spraying. However, there are still challenges such as high cost, low degree of agricultural mechanization, laws and regulations and privacy protection. Future development trends include cost reduction, improvement of data processing and analysis technology, integration with other agricultural technologies, autonomous flight and intelligent technology. The application potential of UAV is huge, which can improve agricultural efficiency and quality, and achieve sustainable development. UAV technology should be popularized and applied, and relevant problems should be solved.

Design of a landing gear system for UAV in complex ground environments

The current UAV technology is developing rapidly and there are various UAV concepts such as hybrid vertical take-off and landing (VTOL) UAV in addition to traditional UAV. The main solution direction of these UAV concepts lies in the compatibility of vertical take-off and landing and long endurance and high load. However, in some complex environments, such as post-earthquake debris, or uneven mountainous terrain, UAV take-off and landing still has great difficulties. Therefore, this paper designs a landing gear that can adapt to complex ground conditions and its corresponding landing algorithm based on DJI's publicly available UAV model. First, a movable landing gear controlled by a linear actuator is used to adapt to a variety of complex terrains. Then, we designed a UAV foot similar to a 3D carved needle with oil-hydraulic cushioning to make it fit more to the complex ground and stand stably on the ground. Then, we obtained the mathematical relationship between the linear actuator and the landing gear moving end based on the simplified mathematical model of the UAV model, and applied it in the subsequent landing algorithm. And then, in detecting the landing point, we use the column coordinate system for coordinate simulation while the algorithm written in C language to get the most suitable UAV landing point. Finally, we have simulated the UAV landing using MATLAB and the results show the high engineering value.

Analysis of development factors of the unmanned aerial vehicles market in the agricultural industry

Current trends in the world economic system require developing new effi cient measures of support for agricultural industry based on the modern methods of production technologization. This is also indicated by the priorities established within the Russian Federation Food Security Doctrine, which presuppose formation of high-performance industries in the agricultural sectors. Thus, current economic conditions and scientifi c and technological development trends determine the causes and form the prerequisites for active development of the UAV market and expansion of the range of their use in various sectors of economy including agriculture. In this context, the author analysed development factors of the UAV market and evaluated the degree of impact of these factors. The hypothesis has been set up that current economic conditions and scientifi c and technological development trends determine the causes and form the prerequisites for active development of the UAV and innovative technologies market and expansion of the range of their use in various sectors of economy including agriculture. The empirical basis of the study includes expert assessments, regulatory legal acts in the sphere of development of agricultural UAV market as well as offi cial data from statistics agencies. Matrix factor analysis and ranking of expert assessments have been applied as the basic method. The author gives a ranked quantitative assessment of the factor impact on the UAV market including ranking of expert assessments. The author carries out detailed elaboration of expert assessments and on this basis identifi es major problems arising under the infl uence of heterogeneous factors of mega-, macro-, meso- and microenvironment on the UAV market. Thus, at the macrolevel signifi cant factors include climate change, rise in natural disasters, and need for satisfying the increasing demand on food products. Considering the factors revealed it is essential to establish strategic priorities for the development of the country’s agriculture.

Measurement of near-coastal wind field characteristics based on UAV

Wind load is an important factor controlling structural design, and the study of wind field characteristics has been area of interest for many scholars. The bulky nature, high cost and immobility of the experimental equipment are the problems existing in the actual experiment. To study the wind field over a large area, this paper proposes a new type of experimental equipment based on UAV technology. This paper first demonstrates the feasibility of the new equipment by comparing the experimental data with the empirical value. In this paper, the dimensionless method is introduced to effectively solve the problem that traditional instruments cannot obtain wind field data from multiple measuring points at the same time, and the timeliness error caused by different measured times. In order to further explore whether the wind fields of the sea and the beach are significantly different, the profile fitting of the two locations is carried out respectively, and the roughness index is obtained. The analysis results show that when there are only a few trees or buildings with low heights on the coast, the impact on the coastal wind field is small.

Uav intelligent recognition and tracking method based on deep learning

In the era of artificial intelligence, the traditional industry has undergone great changes under the guidance of modern technology theory. Enterprises or institutions in various fields have begun to develop and innovate UAV equipment with the application value of The Times with the help of technologies such as deep learning or machine vision to meet the increasing business processing needs. Since the unmanned aerial vehicle will transmit the captured images to the ground, and the encoding and decoding during transmission will affect the real-time performance of the overall system, scientific researchers have proposed an intelligent recognition and tracking method for unmanned aerial vehicles, in order to improve the work efficiency and quality of the system equipment. On the basis of understanding the current research status of UAV and target recognition technology, this paper mainly studies the UAV target recognition algorithm with deep learning as the core according to the intelligent UAV platform design scheme. The final result shows that the deep learning recognition algorithm can further improve the efficiency of UAV intelligent recognition and tracking.