IEEE Transactions on Instrumentation and Measurement最新文献

{"title":"CRKD-YOLO: Cross-Resolution Knowledge Distillation for Low-Resolution Remote Sensing Image Object Detection","authors":"Xiaochen Huang;Qizhi Teng;Hong Yang;Xiaohai He;Linbo Qing;Pingyu Wang;Honggang Chen","doi":"10.1109/TIM.2025.3559616","DOIUrl":"https://doi.org/10.1109/TIM.2025.3559616","url":null,"abstract":"The majority of advanced remote sensing object detection technologies excel in accurately detecting objects from high-resolution images. However, in practical scenarios, it is often necessary to detect objects in images of varying resolutions due to differences in imaging equipment. When dealing with lower-resolution images, the limited detailed information and blurry boundaries lead to a noticeable decrease in detection accuracy. To address this problem, we propose an efficient object detection method for low-resolution remote sensing images based on the YOLO detector, named CRKD-YOLO. The method constructs a cross-resolution knowledge distillation (CRKD) framework to resolve the issue of feature mismatch, enabling the model with low-resolution inputs to learn more refined feature representations from high-resolution images. Furthermore, to effectively leverage the limited detailed information in low-resolution images, we propose the backbone augment feature pyramid network (BAFPN). It enhances detection accuracy for low-resolution remote sensing images while making the model more lightweight. Massive experiments on DOTA, DIOR, NWPU VHR-10, DroneVehicle, and VEDAI demonstrate that our CRKD-YOLO achieves significant improvements, even achieving higher accuracy compare to training and testing high-resolution images with baseline. Our code is published at <uri>https://github.com/Jianfantasy/CRKD-YOLO</uri>","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-17"},"PeriodicalIF":5.6,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143865294","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Implementation of Low-Cost High-Performance Generators for Testing the Harmonic Measurement Accuracy of Instrument Transformers","authors":"Adriano Demetrio;Marco Faifer;Christian Laurano;Roberto Ottoboni;Sergio Toscani","doi":"10.1109/TIM.2025.3554910","DOIUrl":"https://doi.org/10.1109/TIM.2025.3554910","url":null,"abstract":"A proper characterization of instrument transformers requires waveform generators able to apply realistic periodic voltages and currents, resembling those typically found in distribution grids. This article proposes a simple approach for dramatically improving the performance of generators based on the usual, low-cost architecture consisting of a power amplifier and a coupling transformer, which enables reaching the required voltage and current levels. The method is based on iterative frequency-domain error feedback, with feedback gain set according to a preliminary frequency response measurement of the open-loop generation system. The theoretical analysis demonstrates that the asymptotic generation error depends on the adopted reference transducer, on the disturbance level, but not on the characteristics of the generation system. This feature thus enables reaching high generation accuracy without using an overdesigned coupling transformer. The proposed approach has been adopted for the implementation of a high current and a medium voltage generator. The experimental results confirm the effectiveness of the frequency-domain feedback method that in both the cases allows for a remarkable accuracy improvement.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-13"},"PeriodicalIF":5.6,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10962374","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143852396","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Robust Capsule Network With Adaptive Fusion of Multiorder Proximity for Intelligent Decoupling of Compound Fault","authors":"Peirong Zhu;Yongzhi Liu;Tianxing Li;Haoran Du;Ting Liu","doi":"10.1109/TIM.2025.3554898","DOIUrl":"https://doi.org/10.1109/TIM.2025.3554898","url":null,"abstract":"With the advancement of sensor acquisition technology and deep learning algorithms, intelligent fault diagnosis based on equipment operation data has achieved significant progress in the industrial field. However, existing deep learning methods are only aimed at recognizing a single fault, ignoring the concurrence and coupling of various types of faults in industrial scenarios. The presence of compound faults leads to an exponential increase in the number of original fault modes, posing a major challenge in fault diagnosis. To solve this issue, this article proposes a zero-shot compound fault intelligent decoupling method based on a capsule network under the framework of adaptive fusion of multiorder proximity (AFMP) and generalized sparse norm. First, the capsule network with the ability to be sensitive to spatial features is utilized to build an intelligent decoupling model. Subsequently, a dynamic routing scheme with AFMP using Cauchy graph embedding is designed for learning mutual information of both local and global aspects of overlapping features of compound fault, which improves the representation learning ability of the decoupling model. Finally, the generalized sparse <inline-formula> <tex-math>${l_{p}}/{l_{q}}$ </tex-math></inline-formula> norm is introduced to redesign the probabilistic output function for compound fault decoupling, which improves the decoupling generalization and robustness of the model to unknown compound faults under the training using only single-fault samples. To verify the effectiveness of the proposed method, it was validated on a self-made airborne fuel pump (AFP) experimental platform. Extensive results show that our proposed method reaches an optimal average accuracy of 99.66% and 93.4% for decoupling compound fault under constant and varying operating conditions, respectively, without any compound fault samples involved in the model training process and outperforms a series of existing state-of-the-art models.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-15"},"PeriodicalIF":5.6,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143856265","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development of a Vision-Based Ground Target Localization System for Flapping-Wing Flying Robots","authors":"Shengnan Liu;Qiang Fu;Xiaoyang Wu;Wei He","doi":"10.1109/TIM.2025.3559574","DOIUrl":"https://doi.org/10.1109/TIM.2025.3559574","url":null,"abstract":"Implementing ground target localization remains difficult for a flapping-wing flying robot (FWFR) owing to its intrinsically periodic flapping motion and low load capacity. In this article, a vision-based ground target localization system is developed for FWFRs by using two different focal length cameras. First, to counteract the image jitter during the flight of FWFRs, we design a lightweight camera stabilizer based on the motion characteristics of FWFRs and adopt the active disturbance rejection control (ADRC) method rather than the traditional PID control method to obtain smoother aerial videos. Second, a dual-camera system consisting of long- and short-focal-length cameras is designed to eliminate the impact of flight altitude on target detection performance. We combine the digital zoom algorithm with the dual-camera system and propose an improved target detection algorithm based on YOLOv8, which successfully detects ground targets captured by the dual-camera system at different flight altitudes. Then, the latitude and longitude coordinates of the ground target are estimated by fusing the information from cameras and other onboard sensors. Finally, extensive flight experiments carried out using our self-developed FWFR named USTB-Hawk demonstrate the effectiveness of the designed ground target localization system. Our experimental results show that at a flight altitude of 100 m, the average localization error is 4.4 m. At a flight altitude of 300 m, the average localization error is 6.0 m. This provides insights into performing vision-based ground target localization tasks for the FWFR.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-13"},"PeriodicalIF":5.6,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143875181","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"JDCBL: A Joint Dynamic Contexts and Balanced Learning Framework for Real-World Fabric Defect Detection","authors":"Ningli An;Liangliang Wan","doi":"10.1109/TIM.2025.3558828","DOIUrl":"https://doi.org/10.1109/TIM.2025.3558828","url":null,"abstract":"Fabric defect detection is crucially essential for maintaining fabric quality. Existing intelligent detection systems, however, often fall short in real scenarios due to significant variations in defect sizes, which typically manifest as small and elongated structures. In addition, insufficient samples in real scenarios result in a long-tail distribution in training data, posing challenges to the balanced data prerequisites of having collecting balanced data for deep learning methods. This article introduces the joint dynamic contexts and balanced learning (JDCBL) framework to cope with these challenges. Within this framework, contextual transformer block (CoTBlock) is designed to capture both global and fine-grained features of fabric defects simultaneously, which forms contextual transformer network (CoTNet). This modification boosts the interactivity of contextual information between input keys and effectively enhances self-attention learning. Additionally, class average and class complement strategies balanced contrastive loss (BCL) amplify the exposure of tail-class defects so they can be accurately identified. Extensive validation on the Tianchi fabric dataset demonstrates that the JDCBL framework surpasses other leading methods in multiple defect detection tasks, achieving a MAP50 of 39.3%—over 5% higher than competing methods.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-12"},"PeriodicalIF":5.6,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143852440","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Decoupled Parameter Estimation of Small AAV Rotors Using FMCW Radar","authors":"Zhiyong Xu;Chen Chen;Sirui Tian;Min Deng;Zhao Zhao","doi":"10.1109/TIM.2025.3559162","DOIUrl":"https://doi.org/10.1109/TIM.2025.3559162","url":null,"abstract":"The parameter estimation of the rotors of autonomous aerial vehicles (AAVs) based on inverse synthetic aperture radar (ISAR) has great help to the detection and recognition of the AAVs. When facing undersampling in azimuth, most of the existing algorithms have the problem of limited parameter estimation accuracy and high algorithm complexity, and the performance of these algorithms often greatly decreases. In this study, a novel and efficient parameter estimation method of the AAV rotors is proposed. The method first realizes the parameter decoupling, which simplifies the complex multiparameter searching problem into three 1-D parameter estimation problems. More importantly, the proposed strategy has robust performance under low pulse repetition frequency (PRF) conditions, which surpass the existing ISAR-based methods. The scattering point distribution of the rotor blade is recovered with the orthogonal matching pursuit (OMP) method, leading to an accurate estimate of the blade length when facing low PRFs. Experiments on simulated and measured data have shown that the proposed method is effective under different undersampling, signal-to-noise ratio (SNR) and phase noise cases. The rotor parameters are estimated with sufficient accuracy and the running time is only about 10% of the conventional traversal algorithms.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-12"},"PeriodicalIF":5.6,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143875179","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Transfer Learning-Based Multimodal Feature Fusion Model for Bearing Fault Diagnosis","authors":"Honggui Han;Yuan Meng;Xiaolong Wu;Xin Li;Junfei Qiao","doi":"10.1109/TIM.2025.3558745","DOIUrl":"https://doi.org/10.1109/TIM.2025.3558745","url":null,"abstract":"Fault diagnosis based on single-modal features struggles to capture the coupling relationship between multiple fault factors, resulting in inferior diagnosis accuracy. To address this problem, a transfer learning-based multimodal feature fusion (TL-MMFF) model is proposed for fault diagnosis. First, a continuous wavelet transform (CWT)-based modal expression method is employed to transform raw vibration signals into time-frequency representations. Then, this high-resolution time-frequency modal can be utilized to capture transient vibration and energy changes in nonstationary signals. Second, a multimodal feature fusion strategy is proposed, which designs learnable parameters to dynamically weight the time-domain features of torque and the time-frequency features of vibration signals. This adaptive weighting strategy optimizes the fusion process based on the correlation of different modal feature sets, thereby enhancing the ability to describe fault characteristics. Third, a maximum mean discrepancy (MMD)-based transfer learning (TL) algorithm is designed to reduce the distribution differences between fused features under different operating conditions. Then, the model can identify fault characteristics across varying operating conditions. Finally, experiments on the Paderborn University dataset demonstrate that TL-MMFF achieves 99.1% accuracy and converges 30% faster than single-modal methods. These results validate the effectiveness of the model in integrating multimodal data and generalizing across domains.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-13"},"PeriodicalIF":5.6,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143871042","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MixDual-Tuning: Improved Fine-Tuning for Cross-Subject Few-Shot Motor Imagery Classification","authors":"Xun Song;Xinhui Li;Cunhang Fan;Zhen Chen;Hongyu Zhang;Xu Zhang;Fan Li;Zhao Lv","doi":"10.1109/TIM.2025.3556908","DOIUrl":"https://doi.org/10.1109/TIM.2025.3556908","url":null,"abstract":"Motor imagery (MI) brain-computer interfaces (BCIs) face challenges posed by individual differences, and models trained on existing subjects are difficult to apply directly to target subjects. Although transfer-learning-based approaches can help alleviate this problem, they require a large amount of target data for model fine-tuning, which leads to a heavy data collection burden and causes mental fatigue in subjects. Recently, several few-shot learning-based approaches have been applied to MI-BCI, achieving promising performance with a small amount of data on target subjects. However, most existing techniques ignore the overfitting and domain bias issues associated with limited data. To address these challenges, we propose the MixDual-Tuning method, a novel fine-tuning method that combines data augmentation with an improved dual-component loss function. In detail, synthetic data are first generated through augmentation and then combined with target domain data to increase data volume and diversity, reducing overfitting. Moreover, a dual-component loss function is applied to encourage domain-invariant feature learning, by combining cross-entropy loss for classification and optimized MMD loss for domain alignment. We also introduce and enhance a flooding regularization technique to prevent overfitting by stabilizing the training loss. We evaluate the effectiveness of MixDual-Tuning on three publicly available MI-BCI datasets, including competition IV-2a, IV-2b, and our dataset. Extensive experiments demonstrate that MixDual-Tuning consistently surpasses both baseline models and recent few-shot learning approaches, verifying the effectiveness of the proposed method.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-11"},"PeriodicalIF":5.6,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143871121","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Distributed Cooperative Localization for Unmanned Systems Using UWB/INS Integration in GNSS-Denied Environments","authors":"Tuan Li;Xiaoyang Yu;Qiufang Lin;Yuezu Lv;Guanghui Wen;Chuang Shi","doi":"10.1109/TIM.2025.3559161","DOIUrl":"https://doi.org/10.1109/TIM.2025.3559161","url":null,"abstract":"In unmanned systems, integration of inertial measurement unit (IMU) with global navigation satellite systems (GNSSs) provides accurate state information when satellite signals are available. However, during GNSS-denied periods, positioning accuracy degrades rapidly due to the accumulating errors of the inertial navigation system (INS). To improve positioning accuracy in such environments, we propose a distributed cooperative localization (CL) method that leverages relative distance measurements between unmanned systems to mitigate cumulative positioning errors of INS. We first analyze how the cross-covariance matrix and the number of measurements in centralized CL systems, based on extended Kalman filter (EKF), impact positioning accuracy. Our theoretical analysis reveals that the cross-covariance matrix plays a key role in determining localization accuracy, and confirms that propagating the covariance matrix of its own state among individual systems within a cluster is feasible. Based on these insights, we develop a distributed CL algorithm that maintains the cross-covariance matrix while using only a subset of relative distance measurements. The performance of the proposed algorithm is validated through theoretical analysis and field experiments. The results demonstrate that: 1) broadcasting a vehicle’s covariance matrix within a cluster is feasible; 2) compared to INS-only solution, the distributed CL method we proposed reduces root-mean-square error (RMSE) by approximately 50%, with positioning accuracy approaching to that of the centralized CL results; and 3) incorporating known reference point within a cluster can constrain error drift.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-13"},"PeriodicalIF":5.6,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143856327","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"DGDOT-Net: A Deep Generative Model With Attention Fusion for Enhanced High-Density Diffuse Optical Tomography","authors":"Di Wu;Meiyun Xia;Deyu Li;Chuanxin M. Niu;Daifa Wang","doi":"10.1109/TIM.2025.3554283","DOIUrl":"https://doi.org/10.1109/TIM.2025.3554283","url":null,"abstract":"Functional near-infrared spectroscopy (fNIRS) noninvasively evaluates the optical properties of target tissues to monitor functional changes. High-density diffuse optical tomography (HD-DOT) based on this technology enables high-resolution 3-D reconstruction. However, the strong scattering of photons by brain tissue limits the ability of detected signals to accurately reflect changes in brain function, reducing both the accuracy and 3-D resolution of fNIRS-based reconstructions. This article introduces a deep generative model, DGDOT-Net, which incorporates an attention fusion mechanism to enhance the imaging resolution and robustness. The model first decouples key features in the inverse mapping process between observed signals and reconstructed results, leveraging the conditional variational autoencoder (CVAE) architecture to model the probability distribution in latent space and regulate the reconstruction outcome. In addition, a depth-aware attention mechanism embedded within the encoder and decoder extracts effective features from the progressive encoding process, improving learning efficiency. This study first demonstrates the superior reconstruction performance of the model through a series of numerical simulation experiments and evaluates its robustness under low signal-to-noise ratios and varying medium conditions. Specifically, the average values of structural similarity index (SSIM), peak signal-to-noise ratio (PSNR), mean absolute error (MAE), contrast-to-noise ratio (CNR), R, and Jaccard Index achieved by DGDOT-Net on simulated data are 0.83, 21.03 dB, <inline-formula> <tex-math>$2.75times 10 ^{mathrm {-3}}$ </tex-math></inline-formula>, 7.10, 0.45, and 0.73, respectively. Subsequently, physical phantom data collected using a locally developed prototype system are tested, yielding average metric values of 0.87, 18.65 dB, <inline-formula> <tex-math>$15.32times 10 ^{mathrm {-3}}$ </tex-math></inline-formula>, 9.61, 0.79, and 0.92, respectively. Furthermore, DGDOT-Net demonstrates the ability to reconstruct optical properties at a depth of 1.5 cm with a spatial resolution of 1 cm. The experimental results confirm that the proposed model enhances the 3-D reconstruction of brain functions using fNIRS, advancing the clinical applicability of related technologies.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-20"},"PeriodicalIF":5.6,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143848862","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}