Journal Information
EISSN : 2504-446X
Published by: MDPI (10.3390)
Total articles ≅ 508
Current Coverage
DOAJ
SCOPUS
SCIE
INSPEC
Archived in
SHERPA/ROMEO
Filter:

Latest articles in this journal

Published: 4 July 2022
by MDPI
Abstract:
This article presents a review about Beyond Visual Line Of Sight (BVLOS) operations using unmanned aircraft in forest environments. Forest environments present unique challenges for unmanned aircraft operations due to the presence of trees as obstacles, hilly terrain, and remote areas. BVLOS operations help overcome some of these unique challenges; however, these are not widespread due to a number of technical, operational, and regulatory considerations. To help progress the application of BVLOS unmanned aircraft operations in forest environments, this article reviews the latest literature, practices, and regulations, as well as incorporates the practical experience of the authors. The unique characteristics of the operating environment are addressed alongside a clear argument as to how BVLOS operations can help overcome key challenges. The international regulatory environment is appraised with regard to BVLOS operations, highlighting differences between countries, despite commonalities in the considerations that they take into account. After addressing these points, technological, operational, and other considerations are presented and may be taken into account when taking a risk-based approach to BVLOS operations, with gaps for future research to address clearly highlighted. In totality, this article provides a practical understanding of how BVLOS unmanned aircraft operations can be done in forest environments, as well as provides a basis for future research into the topic area.
Published: 4 July 2022
by MDPI
Abstract:
Distributed multi-agent collaborative decision-making technology is the key to general artificial intelligence. This paper takes the self-developed Unity3D collaborative combat environment as the test scenario, setting a task that requires heterogeneous unmanned aerial vehicles (UAVs) to perform a distributed decision-making and complete cooperation task. Aiming at the problem of the traditional proximal policy optimization (PPO) algorithm’s poor performance in the field of complex multi-agent collaboration scenarios based on the distributed training framework Ray, the Critic network in the PPO algorithm is improved to learn a centralized value function, and the muti-agent proximal policy optimization (MAPPO) algorithm is proposed. At the same time, the inheritance training method based on course learning is adopted to improve the generalization performance of the algorithm. In the experiment, MAPPO can obtain the highest average accumulate reward compared with other algorithms and can complete the task goal with the fewest steps after convergence, which fully demonstrates that the MAPPO algorithm outperforms the state-of-the-art.
Published: 1 July 2022
by MDPI
Abstract:
Although drone thermography is increasingly applied as an archaeological remote sensing tool in the last few years, the technique and methods are still relatively under investigated. No doubt there are successes in positive identification of buried archaeology, and the prospection technique has clear complementary value. Nevertheless, there are also instances where thermograms did not reveal present shallow buried architectural features which had been clearly identified by, for example, ground-penetrating radar. The other way around, there are cases where the technique was able to pick up a signals of buried archaeology at a time of day that is supposed to be very unfavorable for thermographic recording. The main issue here is that the exact factors determining the potential for tracing thermal signatures of anthropomorphic interventions in the soil are many, and their effect, context, and interaction under investigated. This paper deals with a systematic application of drone thermography on two archaeological sites in different soils and climates, one in The Netherlands, and one in Italy, to investigate important variables that can make the prospection technique effective.
Published: 29 June 2022
by MDPI
Abstract:
This study presents a novel distributed behavior model for multi-agent unmanned aerial vehicles (UAVs) based on the entropy of the system. In the developed distributed behavior model, when the entropy of the system is high, the UAVs get closer to reduce the overall entropy; this is called the grouping phase. If the entropy is less than the predefined threshold, then the UAVs switch to the mission phase and proceed to a global goal. Computer simulations are performed in AirSim, an open-source, cross-platform simulator. Comprehensive parameter analysis is performed, and parameters with the best results are implemented in multiple-waypoint navigation experiments. The results show the feasibility of the concept and the effectiveness of the distributed behavior model for multi-agent UAVs.
Published: 29 June 2022
by MDPI
Abstract:
Dealing with the trade-off challenge between computation speed and path quality has been a high-priority research area in the robotic path planning field during the last few years. Obtaining a shorter optimized path requires additional processing since iterative algorithms are adopted to keep enhancing the final optimized path. Therefore, it is a challenging problem to obtain an optimized path in a real-time manner. However, this trade-off problem becomes more challenging when planning a path for an Unmanned Aerial Vehicle (UAV) system since they operate in 3D environments. A 3D map will naturally have more data to be processed compared to a 2D map and thus, processing becomes more expensive and time-consuming. This paper proposes a new 3D path planning technique named the Sobel Potential Field (SPF) technique to deal effectively with the swiftness-quality trade-off. The rationale of the proposed SPF technique is to minimize the processing of potential field methods. Instead of applying the potential field analysis on the whole 3D map which could be a very expensive operation, the proposed SPF technique will tend to focus on obstacle areas. This is done by adopting the Sobel edge detection technique to detect the 3D edges of obstacles. These edges will be the sources of the repulsive forces while the goal point will be emitting an attractive force. Next, a proposed objective function models the strength of the attractive and repulsive forces differently to have various influences on each point on the map. This objective function is then optimized using Particle Swarm Optimization (PSO) to find an obstacle-free path to the destination. Finally, the PSO-based path is optimized further by finding linear shortcuts in the path. Testbed experimental results have proven the effectiveness of the proposed SPF technique and showed superior performance over other meta-heuristic optimization techniques, as well as popular path planning techniques such as A* and PRM.
Published: 27 June 2022
by MDPI
Abstract:
The use of drones in various applications has now increased, and their popularity among the general public has increased. As a result, the possibility of their misuse and their unauthorized intrusion into important places such as airports and power plants are increasing, threatening public safety. For this reason, accurate and rapid recognition of their types is very important to prevent their misuse and the security problems caused by unauthorized access to them. Performing this operation in visible images is always associated with challenges, such as the small size of the drone, confusion with birds, the presence of hidden areas, and crowded backgrounds. In this paper, a novel and accurate technique with a change in the YOLOv4 network is presented to recognize four types of drones (multirotors, fixed-wing, helicopters, and VTOLs) and to distinguish them from birds using a set of 26,000 visible images. In this network, more precise and detailed semantic features were extracted by changing the number of convolutional layers. The performance of the basic YOLOv4 network was also evaluated on the same dataset, and the proposed model performed better than the basic network in solving the challenges. Compared to the basic YOLOv4 network, the proposed model provides better performance in solving challenges. Additionally, it can perform automated vision-based recognition with a loss of 0.58 in the training phase and 83% F1-score, 83% accuracy, 83% mean Average Precision (mAP), and 84% Intersection over Union (IoU) in the testing phase. These results represent a slight improvement of 4% in these evaluation criteria over the YOLOv4 basic model.
Published: 27 June 2022
by MDPI
Abstract:
In this paper, we consider the difference in the abstraction level of features extracted by different perceptual layers and use a weighted perceptual loss-based generative adversarial network to deblur the UAV images, which removes the blur and restores the texture details of the images well. The perceptual loss is used as an objective evaluation index for training process monitoring and model selection, which eliminates the need for extensive manual comparison of the deblurring effect and facilitates model selection. The UNet jump connection structure facilitates the transfer of features across layers in the network, reduces the learning difficulty of the generator, and improves the stability of adversarial training.
Published: 27 June 2022
by MDPI
Abstract:
Due to the nonlinear and asymmetric input constraints of the fixed-wing UAVs, it is a challenging task to design controllers for the fixed-wing UAV formation control. Distance-based formation control does not require global positions as well as the alignment of coordinates, which brings in great convenience for designing a distributed control law. Motivated by the facts mentioned above, in this paper, the problem of distance-based formation of fixed-wing UAVs with input constraints is studied. A low-gain formation controller, which is a generalized gradient controller of the potential function, is proposed. The desired formation can be achieved by the designed controller under the input constraints of the fixed-wing UAVs with proven stability. Finally, the effectiveness of the proposed method is verified by the numerical simulation and the semi-physical simulation.
Published: 27 June 2022
by MDPI
Abstract:
Arbitrary-oriented vehicle detection via aerial imagery is essential in remote sensing and computer vision, with various applications in traffic management, disaster monitoring, smart cities, etc. In the last decade, we have seen notable progress in object detection in natural imagery; however, such development has been sluggish for airborne imagery, not only due to large-scale variations and various spins/appearances of instances but also due to the scarcity of the high-quality aerial datasets, which could reflect the complexities and challenges of real-world scenarios. To address this and to improve object detection research in remote sensing, we collected high-resolution images using different drone platforms spanning a large geographic area and introduced a multi-view dataset for vehicle detection in complex scenarios using aerial images (VSAI), featuring arbitrary-oriented views in aerial imagery, consisting of different types of complex real-world scenes. The imagery in our dataset was captured with a wide variety of camera angles, flight heights, times, weather conditions, and illuminations. VSAI contained 49,712 vehicle instances annotated with oriented bounding boxes and arbitrary quadrilateral bounding boxes (47,519 small vehicles and 2193 large vehicles); we also annotated the occlusion rate of the objects to further increase the generalization abilities of object detection networks. We conducted experiments to verify several state-of-the-art algorithms in vehicle detection on VSAI to form a baseline. As per our results, the VSAI dataset largely shows the complexity of the real world and poses significant challenges to existing object detection algorithms. The dataset is publicly available.
Published: 26 June 2022
by MDPI
Abstract:
Traditional forest monitoring has been mainly performed with images or orthoimages from aircraft or satellites. In recent years, the availability of high-resolution 3D data has made it possible to obtain accurate information on canopy size, which has made the topic of canopy 3D growth monitoring timely. In this paper, forest growth pattern was studied based on a canopy point cloud (PC) reconstructed from UAV aerial photogrammetry at a daily interval for a year. Growth curves were acquired based on the canopy 3D area (3DA) calculated from a triangulated 3D mesh. Methods for canopy coverage area (CA), forest coverage rate, and leaf area index (LAI) were proposed and tested. Three spectral vegetation indices, excess green index (ExG), a combination of green indices (COM), and an excess red union excess green index (ExGUExR) were used for the segmentation of trees. The results showed that (1) vegetation areas extracted by ExGUExR were more complete than those extracted by the other two indices; (2) logistic fitting of 3DA and CA yielded S-shaped growth curves, all with correlation R2 > 0.92; (3) 3DA curves represented the growth pattern more accurately than CA curves. Measurement errors and applicability are discussed. In summary, the UAV aerial photogrammetry method was successfully used for daily monitoring and annual growth trend description.
Back to Top Top