Reconnaissance Applications Research Articles

An analytical analysis of vibrational energy harvesting using piezoelectric energy harvester in a fixed wing unmanned aerial vehicle

Fixed-wing unmanned aerial vehicles (UAVs) in the modern day have widespread applications in various reconnaissance, agriculture, military, and civilian operations. Their wide domain of application has placed the development of drone technology at the forefront. The energy harvesting in UAVs seems promising since it not only has the ability to lengthen flying time but also guarantees that no appreciable weight is added. The current article explores the concept of placing piezoelectric material patches on the UAV wing in order to extract energy from the vibration of the wing. This study uses a UAV's fixed NACA 2412 wing to harvest vibration energy by finding the location and placing a piezoelectric patch where it will produce the most voltage. The mode shapes generated are used to get the appropriate location to place the piezoelectric patch. The fixed-wing of a UAV with a piezo patch placed is subjected to a harmonically varying pressure load of 213.934 Pa and the behavior of the wing and its frequency response is analyzed in terms of displacement and voltage as output. The maximum voltage produced by the single piezoelectric patch is found about 0.11 V at the first mode shape of the wing. Embedding and increasing the patch at a location of high displacements help to harvest more energy and can be utilized for other purposes.

Explainable artificial intelligence models for enhancing classification reliability of ground weapon systems

This study focused on the development of a reliable artificial intelligence (AI) model to enhance the classification reliability of ground weapon systems for surveillance and reconnaissance applications. The proposed AI model overcomes the limited data availability of military objects such as tanks, canons, and multiple-launch rockets by leveraging transfer learning and fine-tuning techniques. A comprehensive evaluation of 35 deep learning models using the publicly available Military-Vehicles dataset on Kaggle identified MobileNet as the most suitable model for ground weapon system classification. The selected MobileNet model achieved an average F1 score of 92% when tested on a dataset comprising five types of ground-weapon systems. In addition, the application of the explainable AI technique Grad-CAM provided insights into the decision-making process of the proposed model and verified its reliability. Real-world evaluations using frames extracted from training videos demonstrated promising accuracy for tanks, canons, and multiple-launch rockets. However, challenges related to object occlusion and the absence of target objects in the images were observed, which resulted in misclassifications. Overall, this study contributes to the development of explainable and reliable AI models for enhancing the performance of ground surveillance and reconnaissance systems.

Clustering-Based Multi-Region Coverage-Path Planning of Heterogeneous UAVs

Unmanned aerial vehicles (UAVs) multi-area coverage-path planning has a broad range of applications in agricultural mapping and military reconnaissance. Compared to homogeneous UAVs, heterogeneous UAVs have higher application value due to their superior flexibility and efficiency. Nevertheless, variations in performance parameters among heterogeneous UAVs can significantly amplify computational complexity, posing challenges to solving the multi-region coverage path-planning problem. Consequently, this study studies a clustering-based method to tackle the multi-region coverage path-planning problem of heterogeneous UAVs. First, the constraints necessary during the planning process are analyzed, and a planning formula based on an integer linear programming model is established. Subsequently, this problem is decomposed into regional allocation and visiting order optimization subproblems. This study proposes a novel clustering algorithm that utilizes centroid iteration and spatiotemporal similarity to allocate regions and adopts the nearest-to-end policy to optimize the visiting order. Additionally, a distance-based bilateral shortest-selection strategy is proposed to generate region-scanning trajectories, which serve as trajectory references for real flight. Simulation results in this study prove the effective performance of the proposed clustering algorithm and region-scanning strategy.

Finite-time robust formation control of multiple aerial robotic vehicles with uncertainties and time-varying complex perturbations

Formation control of aerial robotic vehicles (ARVs) has a wide range of applications in the battlefield reconnaissance, medical rescue and load transportation, etc. System parameter variations and complex disturbances are inevitable in the formation control of ARVs, and the effects of them on the formation stability control of ARVs are unnegligible. This paper addresses the robust formation control problem of finite-time leader–follower for multiple ARVs suffering from parameter uncertainties and time-varying perturbations. To ensure the states (e.g. position and velocity) of all followers to converge to leader in finite time, a novel finite-time high-order sliding mode consensus control (HOSMCC) scheme is designed. To better cope with the parameter uncertainties and time-varying disturbances, improve formation control accuracy and achieve the robust formation control of finite-time leader–follower in multiple ARV systems, a new finite time high-order sliding mode formation control (HOSMFC) scheme on the basis of finite-time HOSMCC is proposed. The finite-time stability of multi-ARV formation control system is guaranteed, and the desired formation pattern of multi-ARV systems is achieved using Lyapunov stability theorem. Performance comparisons with proportional differential formation controller (PDFC) and sliding mode formation controller (SMFC) are studied on a four-ARV formation control system. The results validate the designed formation control scheme.

Semantic Attention and Structured Model for Weakly Supervised Instance Segmentation in Optical and SAR Remote Sensing Imagery

Instance segmentation in remote sensing (RS) imagery aims to predict the locations of instances and represent them with pixel-level masks. Thanks to the more accurate pixel-level information for each instance, instance segmentation has enormous potential applications in resource planning, urban surveillance, and military reconnaissance. However, current RS imagery instance segmentation methods mostly follow the fully supervised paradigm, relying on expensive pixel-level labels. Moreover, remote sensing imagery suffers from cluttered backgrounds and significant variations in target scales, making segmentation challenging. To accommodate these limitations, we propose a semantic attention enhancement and structured model-guided multi-scale weakly supervised instance segmentation network (SASM-Net). Building upon the modeling of spatial relationships for weakly supervised instance segmentation, we further design the multi-scale feature extraction module (MSFE module), semantic attention enhancement module (SAE module), and structured model guidance module (SMG module) for SASM-Net to enable a balance between label production costs and visual processing. The MSFE module adopts a hierarchical approach similar to the residual structure to establish equivalent feature scales and to adapt to the significant scale variations of instances in RS imagery. The SAE module is a dual-stream structure with semantic information prediction and attention enhancement streams. It can enhance the network’s activation of instances in the images and reduce cluttered backgrounds’ interference. The SMG module can assist the SAE module in the training process to construct supervision with edge information, which can implicitly lead the model to a representation with structured inductive bias, reducing the impact of the low sensitivity of the model to edge information caused by the lack of fine-grained pixel-level labeling. Experimental results indicate that the proposed SASM-Net is adaptable to optical and synthetic aperture radar (SAR) RS imagery instance segmentation tasks. It accurately predicts instance masks without relying on pixel-level labels, surpassing the segmentation accuracy of all weakly supervised methods. It also shows competitiveness when compared to hybrid and fully supervised paradigms. This research provides a low-cost, high-quality solution for the instance segmentation task in optical and SAR RS imagery.

An Infrared Maritime Small Target Detection Algorithm Based on Semantic, Detail, and Edge Multidimensional Information Fusion

The infrared small target detection technology has a wide range of applications in maritime defense warning and maritime border reconnaissance, especially in the maritime and sky scenes for detecting potential terrorist attacks and monitoring maritime borders. However, due to the weak nature of infrared targets and the presence of background interferences such as wave reflections and islands in maritime scenes, targets are easily submerged in the background, making small infrared targets hard to detect. We propose the multidimensional information fusion network(MIFNet) that can learn more information from limited data and achieve more accurate target segmentation. The multidimensional information fusion module calculates semantic information through the attention mechanism and fuses it with detailed information and edge information, enabling the network to achieve more accurate target position detection and avoid detecting one target as multiple ones, especially in high-precision scenes such as maritime target detection, thus effectively improving the accuracy and reliability of detection. Moreover, experiments on our constructed dataset for small infrared targets in maritime scenes demonstrate that our algorithm has advantages over other state-of-the-art algorithms, with an IoU of 79.09%, nIoU of 79.43%, F1 score of 87.88%, and AuC of 95.96%.

Nadar’s Aerial Photography: Science, Ecstasy, Projection

Félix Nadar is the legendary inventor of aerial photography, the origin of the aerial images which are widely used today in various fields. Focusing on his aerial photography and aerial experience, this article reveals the contradictory tendencies in his aerial view. His view is first of all a question of scientific application. He found in aerial photography applications in military reconnaissance and cartography. But Nadar’s aerial view is also linked with romantic ecstasy, stemming not only from the spectacular perception of looking down from the balloon, but also from the sublime sensation he experiences in the sky. Both these views are possible only because of geometric projection, an intrinsic principle of photography and the human eye. The principle of projection of photography enables reconnaissance of the referent and its cartographic measurement through photographs. Because of the principle of projection, Nadar felt a sense of ecstasy from the balloon. Nadar’s dialectical aerial view is symptomatic of the diverse aspects of aerial viewing that will appear after him, and the functions and geometric properties of his aerial photography go beyond the realm of that genre and can be applied to many other photographic fields.

Contrastive Learning Network Based on Causal Attention for Fine-Grained Ship Classification in Remote Sensing Scenarios

Fine-grained classification of ship targets is an important task in remote sensing, having numerous applications in military reconnaissance and sea surveillance. Due to the influence of various imaging factors, ship targets in remote sensing images have considerable inter-class similarity and intra-class difference, which brings significant challenges to fine-grained classification. In response, we developed a contrastive learning network based on causal attention (C2Net) to improve the model’s fine-grained identification ability from local details. The asynchronous feature learning mode of “decoupling + aggregation” is adopted to reduce the mutual influence between local features and improve the quality of local features. In the decoupling stage, the feature vectors of each part of the ship targets are de-correlated using a decoupling function to prevent feature adhesion. Considering the possibility of false associations between results and features, the decoupled part is designed based on the counterfactual causal attention network to enhance the model’s predictive logic. In the aggregation stage, the local attention weight learned in the decoupling stage is used to carry out feature fusion on the trunk feature weight. Then, the proposed feature re-association module is used to re-associate and integrate the target local information contained in the fusion feature to obtain the target feature vector. Finally, the aggregation function is used to complete the clustering process of the target feature vectors and fine-grained classification is realized. Using two large-scale datasets, the experimental results show that the proposed C2Net method had better fine-grained classification than other methods.

AI-based UAV navigation framework with digital twin technology for mobile target visitation

Unmanned Air Vehicles (UAVs), i.e. drones, have become a key enabler technology of many reconnaissance applications in different fields, such as military, maritime, and transportation. UAVs offer several benefits, such as affordability and flexibility in deployment. However, their limited flight time due to energy consumption is one of the key limitations. Therefore, it is crucial to ensure that UAVs can complete the mission while consuming the least energy possible. In this paper, we propose a novel framework for UAV smart navigation to minimize the time and energy of planning mobile targets visitation. We develop a Deep Reinforcement Learning (DRL) approach to allow the drone to learn the targets’ mobility pattern and build its least energy scanning strategy accordingly. We conduct an initial evaluation of the system and our proposed DRL model policy using simulation. Then, to overcome the time-consuming exploration phase of DRL, we develop a Digital Twin (DT) environment of 3D physics-based simulator, which can be used to train the DRL agent efficiently. We also developed a testbed based on hardware integration with the parrot ANAFI drone to verify the feasibility of the proposed methodology. Our findings confirm that the DRL-based agent can achieve performance close to that of a benchmark policy. Moreover, the testbed experiment validates the practicality of utilizing the DT environment for DRL exploration.

Moving target detection based on background modeling and frame difference

The integration of Unmanned Aerial Vehicle (UAV) technology with moving target detection has diverse applications in military reconnaissance, space remote sensing, and smart cities. Traditional motion-based target detection algorithms offer fast processing speeds but lack accuracy. Deep learning-based algorithms, while accurate for specific targets only, are complex and not suitable for resource-limited UAV platforms and lack real-time performance. Therefore, this study proposes a real-time moving target detection algorithm for UAV platforms based on traditional frame difference algorithm. The purpose of this algorithm is to improve detection accuracy, which has been hindered by the limitations of traditional algorithms caused by camera shake, background changes, and fast-moving targets. The algorithm involves rough background modeling, background updating during subsequent video image sequences, image morphology processing, and background compensation. Experimental results from multiple sets of UAV-borne video data demonstrate the algorithm's high target detection rate, low false alarm rate, and ability to detect moving targets stably in complex environments. The proposed algorithm achieves a speed of 25 FPS and a detection accuracy of 91.8%, meeting real-time and accurate detection requirements.

Strong Observability of LTV Systems With Feedthrough and On-Orbit Reconnaissance and Evasion Applications

This letter considers linear time-varying control systems with feedthrough. A necessary and sufficient condition for strong observability (also known as observability with unknown inputs) is stated in terms of the observability matrix and newly redefined invertibility matrix. This is followed by an observer for pointwise reconstruction of the state from the output and its time derivatives. A variable-time weakly unobservable subspace is then introduced such that the rank test can be recast in terms of this subspace and the kernel of a certain matrix. The subspace characterization leads to results on the existence and construction of unobservable state and control functions. The theoretical results are illustrated in an on-orbit reconnaissance application. Within the context of a time-varying relative orbital motion model, it is shown that strong observability is satisfiable with any constant or bounded feedthrough matrix. In the absence of strong observability, control functions to evade observation are constructed.

IFAA: An Intelligent Framework Aware Algorithm to Determine the Boundary of Area under Attack in Military Surveillance and Reconnaissance WSN

Wireless sensor networks (WSNs) have proven effective in military applications of surveillance and reconnaissance. Sensors capable of detecting pressure, temperature, movement and presence of specific chemicals are deployed in such applications. Traditionally, sensor data is collected and transferred to a centralized high-capacity node or control station. Analysis of data is carried out at such centralized facilities. Information or intelligence gathered from sensor data after analysis is used to generate control and management commands that are relayed back to sensor nodes. The situation is analogous to an actual wartime scenario where soldiers who are on the field are equivalent to the sensors. Soldiers observe and sense the situation and communicate their observations to the decision maker who is stationed in the control tent. On gathering field information, the decision maker analyses the data and arrives at his decision which is again communicated to the soldiers on the field. Soldiers as well as sensors are not placed illogically or randomly but intentionally and strategically. Observations made on the field ultimately affect how the soldiers or sensors continue to function. Intelligence gained on the field ultimately gets used on the field itself. Our attempt is to observe, analyze and apply intelligence on the field itself. This work proposes an intelligent algorithm that is aware of the sensor network topology, analyses sensor data within the network and uses the network framework to arrive at usable intelligence. Locally generated intelligence avoids communication to and from the command/control and adds value to military surveillance and reconnaissance applications of WSN. Intelligent sensor management allows us to use just the necessary number of sensors while saving resources on otherwise redundant expenditure. In the present work we have designed and applied a dynamic boundary computation algorithm to determine the boundary of the area under attack. We have compared the results of simulation experiments incorporating the proposed algorithm against a control experiment without the algorithm.

Martian Combustion-Powered Fixed-Wing UAVs: An Introductory Investigation and Analysis

The Martian topography needs to be investigated in greater detail for human habitations, and this can be accomplished faster using unmanned aerial vehicles (UAVs). In this regard, the RQ-11B Raven appears suitable for remote sensing and topography-mapping applications on Mars, due to its popularity in surveillance and reconnaissance applications on Earth. As a result, this study investigates the flight of this UAV in the Martian atmosphere with the assumptions that it employs an NACA S7012 airfoil and its electric propulsion technology is replaced with a four-stroke oxy-methane fueled Saito FG-11 internal combustion engine (ICE). This ICE is estimated to supply 367.8 W resulting in an engine speed of 6891 revolutions per minute. Based on this speed, the UAV must fly at least 72 m/s (Re = 18,100) at a 5° angle of attack to support flight under calm conditions. To achieve this speed will be difficult; thus, a weather balloon or German V1-style launch system should be employed to launch the UAV successfully. Furthermore, the UAV must operate below 165 m/s (Re = 41,450) to prevent transonic conditions. Finally, the vehicle’s fuel and oxidizer tanks can be refueled using an in situ methane and oxygen production system, enabling its sustainable use on Mars.

Adaptive block backstepping control for a UAV performing lateral maneuvers under lateral c.g. uncertainty

PurposeUnmanned aerial vehicles (UAVs) have wide applications in surveillance and reconnaissance without risking human life. Due to unbalanced payload distribution or in-flight deployment, UAVs may undergo lateral center of gravity (c.g.) variations resulting in an asymmetric dynamic having significant longitudinal and lateral/directional coupling and hence more pronounced nonlinearity. Therefore, automatic control of UAVs becomes extremely difficult when it is forced to perform maneuvers under such imbalance in lateral mass distribution. The purpose of this paper is to design adaptive nonlinear control so that the UAV can perform some useful lateral/directional maneuver under lateral c.g. uncertainty.Design/methodology/approachFirst the nominal lateral/directional dynamics of a fixed-wing UAV is framed into strict feedback form and then the block backstepping approach is used to design the controller to execute horizontal turn and aileron roll maneuvers under no lateral c.g. variation. Thereafter, an adaptive block backstepping controller is designed to adapt to uncertainty in lateral c.g. position considering an approximate model of the asymmetric dynamics. The proposed adaptive scheme is validated against the same two maneuvers as considered for the nominal case.FindingsFirst it is shown that the lateral/directional dynamics of a UAV can be converted to a block strict feedback form for executing some lateral/directional maneuvers. However, it was observed that the maneuver performance suffers significant performance degradation under lateral c.g. variations. To mitigate this issue, a simple and computationally inexpensive adaptive block backstepping scheme is proposed and validated. The adaptation law is further proved to be able to asymptotically estimate the actual c.g. location of the UAV.Practical implicationsThe proposed control scheme allows the UAV to automatically adapt to lateral c.g. variations so that the intended maneuvers are performed without any noticeable loss in maneuver performance.Originality/valueThere are very few works available in the literature that address nonlinear control designs for executing specific lateral/directional maneuvers and, moreover, they consider symmetric UAVs or aircraft only. This paper addresses the practical problem of autonomous maneuvering for UAVs with unbalanced lateral mass distribution leading to shift of c.g. out of its plane of symmetry.

(Retracted) Real-time detection of panoramic multitargets based on machine vision and deep learning

With the continuous development of computer vision technology, moving object detection technology has been paid enough attention and made great progress. Many new methods and new equipment have been developed. As an important part of computer vision, it has important applications in battlefield reconnaissance, video surveillance, image compression and retrieval, human–computer interaction and other research fields, moving object detection and tracking algorithm has always been a research hotspot. We mainly study the panoramic multitarget real-time detection based on machine vision and deep learning. By studying the principle of multitarget real-time detection based on machine vision and deep learning, the panoramic multitarget real-time detection model based on machine vision and deep learning is determined. The principle and correction effect of existing image distortion correction algorithm are analyzed, and the existing problems are summarized. Aiming at the problem that the existing image distortion correction effect is not good, a new method based on machine vision and deep learning is proposed. A real-time panoramic multitarget detection method based on degree learning is proposed. The experimental results show that when the target is moving at medium speed and slow speed, the success rate of tracking is 97% and 95%, respectively; the probability of successful target detection is 100% and 97%, respectively. Experimental results show that the improved method can solve the problem of particle degradation and improve the accuracy of real-time detection.

Concurrent Imaging for TerraSAR-X: Wide-Area Imaging Paired With High-Resolution Capabilities

The concurrent imaging technique enables parallel acquisitions with different beams or modes, e.g., a wide area Stripmap mode with a High-Resolution Spotlight mode. Such a concurrent Stripmap/Spotlight imaging technique is investigated for TerraSAR-X. This technique employs a pulse-to-pulse interleaving scheme to acquire two acquisitions-even of disjunctive areas-at the same time, offering products with different resolution and coverage portfolios. This capability is especially interesting for customers interested in an overview of a larger area but at the same time observing an area of interest with higher resolution, e.g., for infrastructure monitoring or reconnaissance applications. The basic concepts, as well as the driving system parameters, are discussed in detail, together with a coverage analysis revealing the high availability rate of the mode combinations on a global scale. A processing approach re-using a substantial part of the existing infrastructure is described and exemplary acquisitions are shown, together with a detailed performance analysis with respect to resolution and ambiguities.

Engineering 1/f noise in porous silicon thin films for thermal sensing applications

The sensitivity of present-day thermal sensors operating in the long-wave infrared region, a critical range for night vision and other reconnaissance applications, is limited by the 1/ f noise in the thermistor material, while their widespread applicability is restricted due to microfabrication incompatibilities with the standard processing for silicon-based microelectronics. Porous silicon films have a high temperature coefficient of resistance ( TCR ~2–10%K −1 ) and are compatible with microfabrication processes, making them suitable candidates for use in thermal sensors. In this work, the 1 /f noise in porous silicon films in conjunction with TCR , both key parameters for high sensitivity thermal sensors are characterized. The inherent 1 /f noise in these as-fabricated films is very high and increases significantly after exposure to atmosphere (oxidation). However, surface passivation at 600 °C, stabilises the film against atmospheric oxidation and reduces the 1 /f noise level by orders of magnitude. Further, decreasing the porosity of passivated films from 90 to 40% decreased the 1 /f noise by an order of magnitude. High TCR ~6–7%K −1 along with low 1/ f noise constant, K~ 10 - 12 are obtainable with an optimised porosity~60–75% for films surface passivated at 600 °C. These results confirm great promise of porous silicon for low manufacturing cost, and higher sensitivity of the thermal sensors. • As-fabricated porous silicon films have very high 1/ f noise that further increases after exposure to atmosphere due to oxidation. • Surface passivation of porous silicon films reduces 1/ f noise by three orders of magnitude. • 1/ f noise can be further tuned by controlling the porosity: decreasing the porosity from 90 to 40% reduced the 1/ f noise constant by an order. • Higher temperature coefficient of resistance ( TCR ) and lower 1/ f noise constant ( K) important to realize high sensitivity based thermal sensors , can be simultaneously obtained by controlling the fabrication parameters and surface passivating the porous silicon films.

Hybrid energy design for lighter than air systems

Solar Powered Lighter Than Air (LTA) systems find applications in reconnaissance, surveillance, mapping of geographical locations, and environmental monitoring using electronic and communication sensor payloads. The system payload comprises an airborne LiDAR for cloud monitoring and a day-night vision camera for surveillance applications. The primary source of payload power is generated by lightweight and flexible solar photovoltaic modules integrated on the platform’s top hull. The generated solar power is affected by wind velocity disturbances, dynamic cloud footprints, and inherent platform contour. The current objective is to design a hybrid energy scheme for uninterrupted payload operation in the presence of the aforementioned degrading factors. A zero-energy deficit design consisting of a diesel generator as an auxiliary power source has been proposed. The novelty of work lies in considering the combined effect of impact factors in the energy scheme for LTA system applications. Under the given environmental conditions, a power outage of 2.5 hours out of 11 operating hours was observed, which is satisfactorily compensated using the proposed hybrid design. The work presents detailed design of a practical and scalable hybrid system which ensures uninterrupted power under realistic environmental conditions.

Design Criteria of a UAV for ISTAR and Remote Sensing Applications

Unmanned aerial vehicle (UAV) is a multi-role avionic system used for many applications. Photogrammetry is the art and science of locating and measuring the geometric property of spatial objects from photographs captured remotely using sensors from air and space. This paper discusses the design of a new integrated system combining the capability of three different technologies, viz. UAV, photogrammetry and artificial intelligence to carry out intelligence, surveillance, target acquisition and reconnaissance (ISTAR) applications. The design criteria of such a UAV are proposed in the form of min–max conditions of these associated technologies. This paper also discusses the improved photogrammetric applications that can be performed using a UAV and artificial intelligence. The size weight area and power (SWAP) constraints in designing such a UAV for photogrammetric and remote sensing applications are discussed in this paper.

Recent advancements in flapping mechanism and wing design of micro aerial vehicles

Recent studies on understanding of natural flyers have encouraged researchers in development of micro aerial vehicles mimicking birds and insects such as hummingbirds, dragonflies, bats and many more. The vehicles find their applications in reconnaissance and situational awareness in combat field, search and rescue operations, biological and nuclear compromised sites and broadcasting and sports. The focus of this review is to assess recent progress in sub systems of these vehicles including drive mechanisms, actuation mechanisms and wing designs that define the aerodynamics, propulsion, stability, and control of the vehicles. Limited research has been carried out on drive mechanisms capable of producing figure-of-eight wingtip motion contrary to conventional four and five-bar linkage mechanisms along with modified planar and spherical attachments. Motor and piezoelectric actuation mechanisms are being used extensively in these vehicles due to lightweight and power efficiency as compared to non-conventional power sources. Wing shape and rigidity plays a key role in determining the required lift and thrust along with frequency limitations and material constraints. A relatively new field of structural and kinematic optimization for the development of a lightweight flapping vehicle with high endurance capability is also a part of this review. This review has pointed out the research gaps including 3-DoF piezoelectric kinematics, under-actuated mechanisms, structural contact analysis, limited static and dynamic structural analysis, limited fatigue analysis and development of optimization techniques.