Aircraft Classification Research Articles

Multi-reference source aircraft classification network based on unknown class prototype estimation

Abstract Gathering complete aircraft types for close-set tasks is challenging and costly in fine-grained aircraft classification, resulting in encountering unknown class aircraft images in real-world models. To address this problem, we propose a network called a multi-reference source fine-grained aircraft classification network (MRSN) to explicitly and implicitly distinguish known class boundaries and embed unknown classes into the sample space. Specifically, we propose an embedding module (UCPE) to synthesize an unknown class prototype to facilitate the modeling of unknown class data distribution. Besides, we introduce a concatenated multi-reference source module (CMS) that combines distance measurement and probability estimation to classify whether testing images are of unknown class or not. Our proposed MRSN innovatively extends the open-set learning paradigm to the fine-grained classification of geospatial objects rather than simple scene classification. Besides, it shows excellent performance in our constructed dataset, which demonstrates the effectiveness and progressiveness of our method.

Impacts of Backcalculation Software Selection on Airport Flexible Pavements Using the Aircraft Classification Rating/Pavement Classification Rating Method

From November 2024, the aircraft classification rating/pavement classification rating (ACR/PCR) method will be accepted by civil aviation authorities to notify the bearing capacity of airport pavements. Structural characteristics are necessary to calculate the parameters relating to this method, such as the modulus of elasticity which, in the case of existing pavements, can be obtained by backcalculation. However, the results of this procedure may vary depending on the type of method and algorithms used to minimize adjustment errors between the deflection basins adopted in the software. Given the above, this article aims to evaluate whether the choice of backcalculation software affects the resistance classification and admissibility of aircraft on airport pavements using the ACR/PCR method based on the backcalculated elastic moduli. The elastic modulus values of the materials from two runways were backcalculated using BAKFAA, BackCAP, BackMedina, and ELMOD software, and then used to calculate the PCR. The results of the backcalculation showed differences in the elastic moduli obtained by the software, mainly in the subgrade. It was found that the backcalculation software used can limit the type of aircraft and the runway resistance. Therefore, the choice of software for backcalculation of deflection basins can influence the results of the ACR/PCR method because of divergences in determining the resistance of the pavement’s bearing capacity and the admissibility of aircraft operations in the analyzed structures.

Radar detectability of light aircraft micro‐Doppler modulation

AbstractThe critical role of specifying micro‐Doppler mode performance in the modelling and development of modern radar systems is investigated. The authors focus on the detection of micro‐Doppler modulation from light aircraft, analysing data from eight helicopters and nine propeller aircraft. With the growing need for accurate target classification in radar technology, incorporating micro‐Doppler detection metrics into radar performance specifications has become increasingly important. This research offers a novel approach to measuring the detectability of micro‐Doppler modulation relative to returns from the main fuselage. The investigation covers the impacts of various preprocessing techniques, polarisation, and aspect angle on detection capabilities. Findings reveal that, on average, micro‐Doppler modulation from propellers is detectable at distances between 50% and 100% of the range at which the fuselage is detected. For helicopters, this range decreases to between 30% and 80%. Additionally, the study introduces empirically derived statistical models designed to predict micro‐Doppler detection ranges in relation to fuselage returns, enhancing the predictability and specificity of radar system performance. This novel contribution presents a basis for improving radar system specifications, leading ultimately to more predictable and reliable light aircraft classification.

Comparison between ACN–PCN and ACR–PCR for rigid airport pavement with case study

This technical note analyses the new Aircraft Classification Rating–Pavement Classification Rating (ACR–PCR) system with a case study. The literatures and case studies showed that anomalies exist between results of pavement structural evaluation of rigid airport pavements using the pavement thickness design (PTD) method and the Aircraft Classification Number–Pavement Classification Number (ACN–PCN) method. In some situations, the ACN–PCN system provides a conservative assessment of pavement damage which will likely overestimate the load impact of aircraft with complex gear configuration. The International Civil Aviation Organisation (ICAO) introduced a new system, the ACR-PCR, which applies layered elastic analysis to estimate pavement bearing capacity. Using a case study from an airport in Southeast Asia, the anomalies between the designed pavement bearing strength of the PTD method and the ACN–PCN method were discussed. The ACR–PCR system can provide pavement structural evaluation results that are consistent with the PTD results.

Border Defence Mechanism Using Deep Learning Technique

In response to escalating concerns regarding potential border airstrikes, this paper introduces an advanced deep-learning model that utilizes Convolutional Neural Networks (CNNs) for accurate military aircraft classification. Functioning as a pivotal tool for proactive threat assessment, the system works by distinguishing and categorizing six aircraft types with exceptional accuracy, facilitating real-time analysis, and aiding with the development of responsive counter-attack strategies. The multi-stage process begins with the comprehensive collection of diverse datasets, followed by the extraction of spatial features from images using CNNs. Optimization techniques fine-tune model parameters to ensure optimal performance. This approach significantly strengthens national security by leveraging advanced technology for proactive threat mitigation. Emphasizing this paper’s vital role in enhancing border defence capabilities, it highlights the model’s capacity to navigate evolving security challenges through improved awareness of the environment. The model's ability to adapt to real-world conditions, including variations in lighting, weather, and terrain, demonstrates its practical applicability. For seamless implementation, the deep learning model is deployed using Django and SQLite as a web page providing an efficient and user-friendly interface This sophisticated deep learning model is fundamental in fortifying national security measures against potential threats along border regions.

Reporting the Bearing Capacity of Airfield Pavements Using PCR Index

Airfield pavements are important assets that have to secure the safe operation of an airport. On this basis, assessing and reporting the bearing capacity of an airfield runway pavement is a critical task. Recently, the Aircraft Classification Rating-Pavement Classification Rating (ACR-PCR) system has been introduced, which uses the PCR index for expressing the bearing capacity of an airfield pavement. In order to accurately determine PCR, the mechanical characteristics and the thicknesses of the individual layers of a pavement are required. For this purpose, it is not seldom that in the absence of resources dedicated to detailed pavement evaluation procedures, assumptions for the material characteristics of the pavement considering typical materials may be made, while pavement thicknesses may be derived by pavement design records. The present paper highlights the importance of using Non-Destructive Testing (NDT) for accurately assessing the in-situ condition of a flexible runway pavement and determining the PCR index. In order to achieve the goal of the investigation, measurements were performed along the flexible pavement of an airport runway. In addition, the paper focuses on the impact of the variation of the thickness and of the mechanical characteristics of the asphalt concrete layers on the PCR index and on the interpretation of the results considering the acceptance of aircraft operations by airport authorities.

The research status and technical situation of electric aircraft

Against the backdrop of escalating global carbon emissions, the emergence of electric aircraft has provided a promising avenue for addressing this pressing challenge. This article discusses the development background, definition, classification, and successful trial cases of electric aircraft, and summarizes the advantages of electric aircraft compared with traditional fossil fuel aircraft. Electric aircraft can be classified into different categories based on several factors, including their power system, purposes, and passenger-carrying capabilities. These classifications help in understanding the diverse range of electric aircraft designs and their suitability for specific purposes. And it has the advantages of emission reduction and noise reduction, economic comfort, high efficiency and reliability, which will have a positive impact on the construction of the new urban transportation system in the future. And then this article introduces the current technological development status from three aspects: electric propulsion system, body material and battery. As an environmentally friendly, efficient and sustainable aviation transportation solution, electric aircraft are receiving more and more attention and research.

Adversarial Learning Improves Vision-Based Perception from Drones with Imbalanced Datasets

This work proposes a vision-based perception algorithm that combines image-processing-based detection and tracking of aerial objects with convolutional neural networks (CNNs) integrated for classification of general aviation aircraft, multirotor small uncrewed aerial systems (SUAS), fixed-wing SUAS, and birds to enable improved onboard avoidance algorithm decision making. Furthermore, we integrate adversarial learning during the training of the CNNs and evaluate performance with class balanced and imbalanced datasets because this maximizes the utility of resource-expensive flight experiments to collect aviation datasets. We compare our proposed CNN with adversarial learning (CNN+ADVL) model with a state-of-the-art CNN as well as a you only look once (YOLO, v4) model retrained (YOLO v4 aircraft) on the same data. The CNN+ADVL trained on the imbalanced dataset achieves the highest 10-fold cross-validation classification accuracy of 76.2% for aircraft and birds for all ranges while achieving 87.0% aircraft classification accuracy, meeting proposed self-assurance separation distances derived from Federal Aviation Administration (FAA) guidelines. In comparison, the CNNs achieved 74.4% 10-fold cross-validation classification accuracy for aircraft and birds as well as 83.4% accuracy for the aircraft, meeting proposed self-assurance separation distances derived from FAA guidelines. Furthermore, we demonstrate that the integration of adversarial learning improves the classification performance for the perception of aerial objects using a class imbalanced dataset.

РЕКОМЕНДАЦІЇ ОРГАНАМ (ПІДРОЗДІЛАМ) ОХОРОНИ ДЕРЖАВНОГО КОРДОНУ ЩОДО ПРОТИДІЇ ПРОТИПРАВНІЙ ДІЯЛЬНОСТІ, ПОВ’ЯЗАНІЙ З НЕЗАКОННИМ ВИКОРИСТАННЯМ ЛІТАЛЬНИХ АПАРАТІВ

In the article, based on the results of the analysis of the recent situation at the state border and the experience of the operational and service activities of the state border protection bodies (units), a conclusion is drawn regarding the activation and expansion of the scale of illegal activities in the border area with the use of various aircraft. According to the results of the analysis of the nature and methods of illegal activity, the structure of channels of smuggling activity, drug trafficking, etc., a classification of aircraft used by violators for the purpose of illegal activity was provided, the technical characteristics and capabilities of aircraft were given, and the possible tactics of actions of violators regarding their use were determined. The above makes it necessary to continue and improve the work of bodies (subdivisions) of the State Border Guard Service of Ukraine to create an effective system for ensuring national security at the state border of Ukraine, taking into account the ways of committing illegal activities, which are constantly being improved. In accordance with the acquired operational experience of the state border protection bodies and units of the State Border Guard Service of Ukraine in countering illegal activities at the state border, options for algorithms for the actions of officials of the body and the state border protection unit when detecting an aircraft violating the state border (a sign of its use) are proposed) and recommendations for combating illegal activities related to their use. The obtained results are aimed at creating an effective system of protecting the state border, introducing new and improving existing methods and ways of countering these offenses. Implementation of the obtained results in the personnel training system will provide an opportunity to expand knowledge on the organization and implementation of measures to combat illegal activities on the state border using aircraft.

ST-Net: Scattering Topology Network for Aircraft Classification in High-Resolution SAR Images

Aircraft classification in synthetic aperture radar (SAR) images plays a considerable role in global region management and surveillance. Recently, deep learning has been applied to solve the classification problem and made significant progress. Due to the imaging variability at different angles and component scattering discreteness in SAR images, previous works have had difficulty in achieving desirable classification results. To address these issues, we study the positional and semantic relationship between the scattering points and propose an innovative scattering topology network (ST-Net) in this article. First, considering the diversity of imaging results caused by different target attitude angles, we extract and transform the scattering cluster centers to update the information of various categories. It can guide the model to strengthen the discriminative features and mitigate the impact of imaging variability on classification performance. Second, a novel scattering topology module (STM) is introduced to model the spatial relationships and semantic information interaction of discrete scattering points. In this process, the topology relations and scattering characteristics are enhanced for further accurate classification. Third, context attention excitation (CAE) is designed to capture significant global and semantic information, which is conducive to suppressing background interference and reducing category confusion. In conclusion, the ST-Net is presented with the SAR imaging mechanism and the topology geometric representation of aircraft. We construct the SAR aircraft category dataset (SAR-ACD) and conduct extensive experiments on it to show the effectiveness of ST-Net, which illustrates that our method achieves superior classification performance.

Runway Pavement Strength Evaluation of Yogyakarta International Airports Depends on ICAO (ACN/PCN) Method with COMFAA 3.0 Software

Yogyakarta International Airport (YIA) is a new airport which serves as an international transit station for Yogyakarta and surrounding areas. The airport was designed to serve large and heavy aircrafts with relatively high frequency. To meet flight safety standards,the runway must be designed to be able withstand the wheel weight of aircrafts to be served by the airport. The study aims to determine the thickness and strength of the runway pavement using the FAA method (Federal Aviation Administration) and COMFAA software. The reference aircraft used was Boeing 747-400ER based on landing wheel configuration. With a subgrade CBR value of 6%, Aircraft Classification Number (ACN) and Pavement Classification Number values of 77.8 and 94.9, respectively, were obtained. PCNvalue higher than ACN value indicates that the pavement structure condition is able to withstand the weight of all aircraft types that are planned to be served by the runway of YIA.

Preliminary evaluation of the ACR-PCR system for reporting the bearing capacity of flexible airfield pavements

Assessing and reporting the bearing capacity of airfield pavements have always been significant tasks in airport engineering, since they support the decision making procedure, which is a core element of Airport Pavement Management Systems (APMS). The official reporting system utilized worldwide during the last four decades is the Aircraft Classification Number – Pavement Classification Number (ACN-PCN) system. However, recently an updated system has been introduced, aiming to overcome the defects of the ACN-PCN system that have been observed during its implementation. This system, referred as the Aircraft Classification Rating – Pavement Classification Rating (ACR-PCR), aims to incorporate the latest advances in airfield pavement design and evaluation and is expected to be fully applicable about 2024. Since this is a transfer period, the current investigation aims to identify any critical aspects arising from the application of the updated system, based on recent Federal Aviation Administration (FAA) airfield pavement evaluation techniques. On this basis, the transferability between the two system is examined, providing useful commentary and remarks for airport authorities, who would like to implement the ACR-PCR method to its full extent. The analysis shows that the transferability between the two systems seems not to be feasible and the implementation of the updated ACR-PCR may present several issues, such as the modification of the existing reported bearing capacity.

Detection with fast feature pyramids and lightweight convolutional neural network: a practical aircraft detector for optical remote images

Aircraft detection in optical remote sensing images has important practical value and attracts great attention in recent years. Due to some clustered backgrounds and various deformations like variability in viewpoints, scales, and directions, it is still a big challenge. In recent years, deep learning, especially convolutional neural networks (CNNs), has achieved remarkable success not only in natural image object detection tasks but also in remote sensing image object detection tasks. However, due to the limitations of the actual application deployment environment only limited processing units and storage space can be used, and most deep learning-based aircraft detection algorithms cannot work well. To this end, a unique aircraft detection framework in optical remote sensing images is proposed, which is based on fast feature pyramids and a specialized lightweight CNN. The algorithm can be summarized into two steps. In the first step, a highly reliable region proposal generation algorithm is designed to predict region proposals. In this phase, a prior square sliding window method and an efficient linear search algorithm are proposed. In the next step, a lightweight CNN is designed to achieve fast and accurate aircraft classification. The proposed model only requires about 237 KB of parameter storage and can infer in a poor central processing unit platform. Comprehensive experiments on three publicly optical remote sensing aircraft datasets demonstrate the superiority and effectiveness of the proposed method. It is also shown that the proposed method is light and fast in comparison with some state-of-the-art methods.

Aircraft Image Recognition Network Based on Hybrid Attention Mechanism.

With the deepening of deep learning research, progress has been made in artificial intelligence. In the process of aircraft classification, the precision rate of aircraft picture recognition based on traditional methods is low due to various types of aircraft, large similarities between different models, and serious texture interference. In this article, the hybrid attention network model (BA-CNN) to implement an aircraft recognition algorithm is proposed to solve the above problems. Using two-channel ResNet-34 as a characteristic extraction function, the depth of network is increased to improve fine-grained characteristic extraction capability without increasing the output characteristic dimension. In the network to introduce a hybrid attention mechanism, respectively, between the residual units of two ResNet-34 channels, channel attention and spatial attention modules are added, more abundant mixed characteristics of attention are obtained, space and characteristics of the local characteristics of the channel response are focused, the characteristics of redundancy are reduced, and the fine-grained characteristics of learning ability are further enhanced. Trained and tested on FGVC-aircraft, a public fine-grained pictures dataset, the recognition precision rate of the BA-CNN networks model reached 89.2%. It can be seen from the experimental results, the recognition precision rate of the original model is improved effectively by using this method, and the recognition precision rate is higher than most of the existing mainstream aircraft recognition ways.

Directional processing in assessment of wind turbine noise

Assessments of wind turbine generator (WTG) noise are required to comply with the US Environmental Agency and local governments and avoid legal action that may result of non-compliant operation. Current methods for WTG noise measurements require the comparison of long-term sound data recorded before and after a WTG installation. These measurements must be conducted during several months for various wind speeds and environmental conditions. Such measurements are complicated due to the presence of multiple sources other than noise not from WTG, including intermittent and fluctuating sources. The application of directional processing can help localize and identify the sources responsible for increased sound levels at any given moment and to estimate the contribution of the other than WTG sources to the whole sound level. Stevens has developed various acoustic systems with several sensors that provide directional processing of sounds that were previously applied for detection, tracking and classification of small boats, divers, small aircraft, and drones. We propose modifications of such systems for directional sensing of WTG noise at long distances in the frequency band relevant for WTG noise measurement.

Classification of Aircraft in Remote Sensing Images Based on Deep Convolutional Neural Networks

Convolutional Neural Network (CNN) is a component of Deep Learning(DL) recently exploited in different fifields. In this work, we improve the performance of multi-label classifification based on CNN for remote sensing images of aircraft types. Intensive preprocessing limits the classifification rate in previous studies. In order to avoid under-fifitting and over-fifitting problems, we optimized the architecture and Network parameters. To validate our method the recent public image dataset called Multi-Type Aircraft Remote Sensing Images (MTARSI) is used. Extensive experiments prove the effectiveness of the proposed method in terms of classifification rate.

Optimizing and Evaluating Swin Transformer for Aircraft Classification: Analysis and Generalizability of the MTARSI Dataset

Aircraft classification via remote sensing images has many commercial and military applications. The Swin-Transformer has shown great promise, recently dominating general purpose image classification benchmarks such as ImageNet. In this manuscript, we test whether the performance of Swin-Transformer on general purpose image classification translate to domain specific aircraft classification using the Multi-Type Aircraft from Remote Sensing Images dataset. We also investigate the effect of training procedure vs. model selection on the validation score. Our carefully trained Swin-Transformer model achieved an impressive 99.4 % validation set accuracy without super-resolution, and 99.5 % with super-resolution. Moreover, the generalization of models trained on the MTARSI dataset to real-world and synthetic aircraft classification is evaluated with some out-of-distribution samples. Our results demonstrated that the lack of complexity and heterogeneity of the MTARSI dataset, and the labelling errors resulted in models which struggle to achieve high accuracy on the adopted test samples despite near perfect validation scores.

Sparse time‐frequency analysis for aircraft target classification with low sampling rate and short observation time

AbstractThe traditional time‐frequency analysis (TFA) techniques are instruments for target classification, which can reflect the feature of the target in the time‐frequency domain. However, it will lead to a serious decrease in the recognition accuracy as the decline of the sampling rate. To ease this problem, in this article, the sparse time‐frequency feature analysis (STFFA) is implemented for aircraft classification, and the genetic algorithm is adopted to solve the sparse problem quickly for saving time. Firstly, the sparse time‐frequency decomposition and recovery signal are obtained by matching pursuit. Then, three novel kinds of aircraft features are extracted from the sparse time‐frequency decomposition, which are sparse recovery time‐frequency entropy (SRTFE), frequency entropy by time, and first‐order sparse time‐frequency moment. Thus, the feature combination modes based on the three features are applied to realize the classification of aircraft and compared with the traditional TFA techniques. Besides, a support vector machine is also used to classify the three kinds of aircraft. The accuracy and efficiency of the STFFA method have been investigated by employing the parametric model data and electromagnetic scattering model simulated trials. Furthermore, in contrast to the traditional TFA instrument, our method can reach a recognition accuracy of more than 90% from the numerical experiment results, which demonstrates that the feature extraction by sparse time‐frequency analysis improves the accuracy of aircraft classification under a low sampling rate.

Multi-scale feature vector reconstruction for aircraft classification using high range resolution radar signatures

High-Resolution Range Profile (HRRP) is effective in various Radar Automatic Target Recognition problems. Multi-scale techniques have been verified in aircraft target HRRP feature enrichment to achieve aircraft recognition performance optimization. The involvement of multiple training-classification procedures in existing multi-scale methods results in tremendous resource consumption which challenges their real-time classification performance and application significance. This paper introduces a novel method that enables multi-scale HRRP features to be manipulated under a single scale for efficiency enhancement. Numerical analysis indicates that multi-scale intensity variations for particular HRRP scatterers could be modeled as Gaussian noises. Linear Discriminant Analysis and Singular Value Decomposition techniques are applied on reconstructed feature vectors for better separability and noise tolerance capability. Experimental results from dynamic aircraft recognition experiments verify the expected efficiency enhancement of the proposed method while maintaining comparable classification accuracy and better noise tolerance performance.

Deep learning for aircraft classification from VHF radar signatures

Radio sources in the Very High Frequency (VHF) band can be seized as opportunity donors in a passive radar configuration such as FM radio stations and VHF omnidirectional range (VOR). A full-wave simulation of three size classes of aeroplanes shows that their bistatic radar cross-section (RCS) are statistically comparable, albeit perform differently in time while the plane is flying. This difference can be exploited to recognize the size of the aeroplanes with respect to these classes. Measurements confirm this possible differentiation between the aeroplanes within the same class. Encouraging initial results were obtained using convolutional or recurrent neural networks to classify aircraft classes, combining simulated bistatic RCS results and real trajectories (collected from automatic dependent surveillance-broadcast data).