{"title":"PRO-CLIP: A CLIP-Based Category Measurement Network Through Prototype and Regularized Optimal Transportation","authors":"He Cao;Yunzhou Zhang;Shangdong Zhu;Lei Wang","doi":"10.1109/TIM.2024.3485403","DOIUrl":"https://doi.org/10.1109/TIM.2024.3485403","url":null,"abstract":"In unstructured environments, robots are likely to encounter desktop objects that they have never seen before. Classifying these objects precisely is a prerequisite for accomplishing object-specific manipulation tasks. However, it is time-consuming to collect large-scale object classification datasets. Inspired by the prompt tuning methods, we propose the PRO-CLIP network, which is a category measurement method for desktop objects. Specifically, PRO-CLIP performs few-shot classification based on the knowledge from pretrained vision-language model (VLM). It utilizes token-level and prompt-level optimal transportations (OTs) to jointly fine-tune the learnable vision-language prompts. For token-level stage, we propose the image patch reweighting (PR) mechanism to make alignments focus on the image patches that are close to the patch prototypes. This allows the patch embeddings have converging category representations, which reduces intraclass differences of visual features. For prompt-level stage, we propose a cascading OT (COT) module to simultaneously consider task-irrelevant knowledge in zero-shot features and task-specific knowledge in few-shot features. Due to the generalization performance of task-irrelevant knowledge, the proposed module achieves feature regularization during OT. Finally, we propose the UP loss to supervise the whole network. It contains unbalanced logit-level consistency losses and visual prototype loss. The logit-level consistency losses are used to make learnable features close to zero-shot features. The prototype loss makes the visual features approach to the corresponding prototypes in distance. We demonstrate the effectiveness of our method by performing few-shot classification experiments on different datasets including desktop objects. The relevant code will be available at \u0000<uri>https://github.com/NeuCV-IRMI/proclip</uri>\u0000.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":null,"pages":null},"PeriodicalIF":5.6,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142595773","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":"Enhanced Azimuth Determination in Drilling via Piecewise Polynomial Fitting and Interpolation","authors":"Tao Guo;Weibin Cheng;Yifei Zhang;Shaobing Hu","doi":"10.1109/TIM.2024.3485449","DOIUrl":"https://doi.org/10.1109/TIM.2024.3485449","url":null,"abstract":"Precise geomagnetic azimuth measurement is essential for automation and intelligent operations in deep-Earth and deep-sea resource exploration. However, during drilling, the main sources of azimuth error include the inherent distortion of triaxial accelerometer and triaxial magnetometer, which requires prior calibration. To address this issue, we focus on static azimuth measurements by analyzing the error propagation relationship, developing a comprehensive error correction model, and proposing a correction method that combines piecewise polynomial fitting and improved interpolation. Finally, the proposed algorithm was experimentally validated using a high-precision inclinometer calibration platform. The experiment compared the error correction effects of the three-step combination method and the proposed method. At a confidence level of \u0000<inline-formula> <tex-math>$p = 0.95$ </tex-math></inline-formula>\u0000 and an axial tilt of \u0000<inline-formula> <tex-math>$I = 0.594$ </tex-math></inline-formula>\u0000, dA decreased from (\u0000<inline-formula> <tex-math>$3.1^{circ }~pm ~0.9^{circ }$ </tex-math></inline-formula>\u0000) to (\u0000<inline-formula> <tex-math>$1.8^{circ }~pm ~0.5^{circ }$ </tex-math></inline-formula>\u0000) and (\u0000<inline-formula> <tex-math>$7times 10^{mathbf {-4}}~pm ~3.7times 10^{mathbf {-2}}$ </tex-math></inline-formula>\u0000) for the traditional and proposed method, respectively. The proposed method demonstrates good performance in azimuth reconstruction and meets the practical engineering demands of azimuth calibration in measurement-while-drilling (MWD) system.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":null,"pages":null},"PeriodicalIF":5.6,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142594976","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 Fabry-Perot Interferometer Sensor Based on a Microsphere Tellurite Fiber Probe for Temperature and Hydraulic Pressure Measurement","authors":"Deyuan Zhong;Yuhan Qu;Qi Wang;Xue Zhou;Xin Yan;Tonglei Cheng","doi":"10.1109/TIM.2024.3485459","DOIUrl":"https://doi.org/10.1109/TIM.2024.3485459","url":null,"abstract":"In this article, a tellurite fiber (TF) was fabricated into a microsphere structure via discharge and sequentially coupled with a multimode fiber and a single-mode fiber (SMF) to form a Fabry-Perot interferometer (FPI) sensor. Changes in temperature and hydraulic pressure cause the microsphere to deform, altering the interference length of reflected lights and establishing their dependency relationship. Experimental results showed that the proposed FPI sensor had temperature sensitivity of 157 pm/°C and hydraulic pressure sensitivity of 11.4 pm/MPa. In addition, repetitive experiments and stability tests were conducted, and the maximum relative standard deviation (RSD) was calculated to be \u0000<inline-formula> <tex-math>$4.34times 10^{-4}$ </tex-math></inline-formula>\u0000 for temperature and \u0000<inline-formula> <tex-math>$4.48times 10^{-4}$ </tex-math></inline-formula>\u0000 for hydra- ulic pressure. This FPI sensor is characterized by small size, compact structure, high sensitivity, long-term accuracy, and good stability and repeatability. All these features render it suitable for real-time temperature and hydraulic pressure monitoring in complex environments. In addition, the tellurite compositions can be adjusted to help adapt the FPI sensor to specific external environmental conditions, which provides valuable theoretical and technical insights for the development of high-performance sensing devices.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":null,"pages":null},"PeriodicalIF":5.6,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142587494","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":"Five-Axis Continuous Sweep Scanning for Curvature Variable Surfaces","authors":"Zherun Li;Yijun Shen;Wenze Zhang;Nuodi Huang;Limin Zhu;Yang Zhang","doi":"10.1109/TIM.2024.3485402","DOIUrl":"https://doi.org/10.1109/TIM.2024.3485402","url":null,"abstract":"Curvature variable surfaces have an extensive application in a multitude of products, such as aeroblades. The aeroblade is an essential part of the aviation industry. The rapid inspection of complex aeroblade is crucial to aviation engine production and maintenance. As an emerging inspection technique, five-axis sweep scanning could achieve higher inspection efficiency compared to the traditional point-by-point probing and line scanning inspection. However, for surfaces with significant curvature variation, such as the blade cross section, the existing five-axis sweep scanning methods require the surface to be separated into patches and conduct inspection on each patch. This type of work pattern necessitates a transition move between each patch, which increases the inspection time and decreases the overall inspection efficiency. In this article, a probe-surface force model is proposed to facilitate the selection of stylus orientations, and an iterative algorithm is introduced to determine the optimal stylus orientations. The guiding curve and the probe head trajectory are synchronized using the dual nonuniform rational basis spline (NURBS) method, which settles the path generation failure at the high curvature region. The proposed method can generate a continuous five-axis sweep scanning path for a blade cross section without transition moves, which drastically boosts the inspection efficiency. Blade cross section inspection experiments are conducted using the proposed method, the patch-wise inspection method, and the line scanning inspection method. The experimental results show that the proposed method maintains the same level of inspection accuracy as both the patch-wise method and the line scanning method. Moreover, the proposed method has achieved a reduction in inspection time of 70.71% over the patch-wise method and 84.98% over the line scanning method, thereby validating its effectiveness.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":null,"pages":null},"PeriodicalIF":5.6,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142595794","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":"Modeling and Characterization of LiDAR Echo-Waveforms in Fog With Experiment Validations","authors":"Ruiqin Yu;Xiaolu Li;Tengfei Bi;Tao Zhang;Zongyu Liu;Landa Gao;Lijun Xu","doi":"10.1109/TIM.2024.3485431","DOIUrl":"https://doi.org/10.1109/TIM.2024.3485431","url":null,"abstract":"Light detection and ranging (LiDAR) generates undesirable clutter signals in fog, impeding target recognition ability. To identify fog clutter waveforms, a comprehensive LiDAR echo waveform generation model is established based on the mechanism of photon random movement, involving factors of environmental particle state, systematic parameters, and target characteristics. The emitted-and-returned photon bundle’s random motions are innovatively described as the superposition of multiple random scattering, and analytical formulations are deduced for photon bundle reception probability, guaranteeing both model accuracy and computational efficiency. Validation experiments are conducted in a large-scale fog chamber. The R-squared values between echo waveforms estimated from our model and measured data achieve \u0000<inline-formula> <tex-math>$0.8307sim 0.9754$ </tex-math></inline-formula>\u0000 for fog clutter, and \u0000<inline-formula> <tex-math>$0.9522sim 0.9778$ </tex-math></inline-formula>\u0000 for target, outperforming the contrasting models. In the simulation model and measurement, fog clutter waveforms exhibit right-skewed asymmetric patterns, enabling their easy differentiation from the Gaussian-distributed target echoes. The presented model and results can be expanded to analyze various atmospheric conditions, broadening the application scenarios of LiDAR.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":null,"pages":null},"PeriodicalIF":5.6,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142595796","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":"DFKD: Dynamic Focused Knowledge Distillation Approach for Insulator Defect Detection","authors":"Bao Liu;Wenqiang Jiang","doi":"10.1109/TIM.2024.3485446","DOIUrl":"https://doi.org/10.1109/TIM.2024.3485446","url":null,"abstract":"Although these methods (e.g., lightweight structural design, model pruning (MP), and model quantization) can reduce the deployment difficulty of deep-learning models in insulator defect (ID) detection, they significantly reduce the detection accuracy. In response to the above issues, this article proposes a dynamic-focused knowledge distillation (DFKD) approach to construct a knowledge transfer path from the large model to the lightweight small model. First, the important sample focusing mechanism introduces dual focus weight factors and adaptive sample matching to encourage the student model to focus on high-quality difficult samples, to reduce the adverse effects of low-quality simple samples. Second, the adversarial training process of the temperature dynamic learning mechanism constructs soft labels of appropriate difficulty based on different stages of distillation training. This helps improve the learning and generalization abilities of the student model toward higher order knowledge. Finally, this article combines the DFKD with the MP to establish an insulator defect detection model [DFKD-MP-You only look once (YOLO)] suitable for edge devices with different computing resources. Experiments show that the DFKD method proposed in this article outperforms existing knowledge distillation (KD) methods in insulator defect detection. Moreover, compared with existing methods (see, e.g., BiFusion-YOLOv3, InsuDet, and ID-YOLO), the DFKD-MP-YOLO not only has a lighter structure, but also achieves higher accuracy and faster speed.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":null,"pages":null},"PeriodicalIF":5.6,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142595911","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}
Xiangyu Meng;Hongyao Tang;Xiaozhou Lü;Yaoguang Shi;Weimin Bao
{"title":"Tactile Electronic Skin With Curved Surface Compensation by In Situ Curvature Self-Sensing","authors":"Xiangyu Meng;Hongyao Tang;Xiaozhou Lü;Yaoguang Shi;Weimin Bao","doi":"10.1109/TIM.2024.3485455","DOIUrl":"https://doi.org/10.1109/TIM.2024.3485455","url":null,"abstract":"Tactile electronic skins with flexibility and good conformability have good application prospects in wearable electronics. However, pseudo-signals are generated when the electronic skin is attached to a curved surface, which affects the pressure detection performance of the electronic skin and limits its application. In this article, we developed a tactile electronic skin that compensates for curved surfaces through in situ curvature sensing and established a pressure, curvature-current model for the compensation of the curvature effect of the electronic skin. The electronic skin can compensate for the measured pressure according to the model and realizes a high-precision sensing capability. It consists of Ag/Ecoflex electrodes, pressure-sensitive materials, and a support layer. The Ag/Ecoflex electrodes with curvature sensing were prepared using screen printing and prestretching. The pressure insensitivity mechanism of the electrodes was analyzed. The effects of curvature on the measurement error of electronic skin were investigated. Experimental results suggest that Ag/Ecoflex electrodes with 80% prestretch have a 21.6/cm−1 curvature sensitivity coefficient and insensitivity to pressure at 0–200 kPa. The measurement errors introduced by curvature can be reduced by compensation. As an application, the skin is attached to the human arm for pressure measurement, through which its great potential is therefore highlighted in curved scenarios such as human prostheses, robots, and flexible airplane skins.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":null,"pages":null},"PeriodicalIF":5.6,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142595103","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":"HGAIQA: A Novel Hand-Geometry-Aware Image Quality Assessment Framework for Contactless Palmprint Recognition","authors":"Chunsheng Zhang;Xu Liang;Dandan Fan;Junan Chen;Bob Zhang;Baoyuan Wu;David Zhang","doi":"10.1109/TIM.2024.3485454","DOIUrl":"https://doi.org/10.1109/TIM.2024.3485454","url":null,"abstract":"Contactless palmprint recognition (PPR) has gained traction due to its convenience and hygienic benefits. However, in real-world scenarios with complex backgrounds and varying hand poses, evaluating image quality to enhance recognition performance remains a significant challenge. To address this, we propose a novel hand-geometry-aware contactless palmprint image quality assessment (HGAIQA) framework. Unlike existing methods that assess only the palmprint region of interest (ROI), our framework evaluates the entire image. First, it employs a high-resolution hand segmentation network and keypoint heatmap module to identify hand region and joint keypoints. Second, it evaluates the palm’s flatness based on geometric features and assesses additional quality attributes such as brightness and sharpness. At last, it determines image quality by analyzing the intraclass and interclass distributions of fused multifeatures. After integrating with subsequent ROI localization and recognition algorithms, experiments show a substantial 21.2% reduction in equal error rate (EER) for PPR on the COEP database by removing the lowest 10% of low-quality images. These results demonstrate the effectiveness of our approach in significantly enhancing PPR performance.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":null,"pages":null},"PeriodicalIF":5.6,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142595838","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":"Fusion Positioning of GNSS-Cellular Signals of Opportunity Under Low Observability","authors":"Tian Jin;Pei Zhang;James Chakwizira;Yuchen Wang","doi":"10.1109/TIM.2024.3485434","DOIUrl":"https://doi.org/10.1109/TIM.2024.3485434","url":null,"abstract":"Cellular signals of opportunity (CSOPs) can be utilized as a backup positioning system for global navigation satellite systems (GNSSs) in urban. Current studies on the fusion of GNSS-CSOPs require sufficient signals. However, in low-observability environments where the total number of visible signals does not exceed four, the existing GNSS-CSOPs pseudorange fusion methods lack robust strategies and have poor performance. In addition, GNSS-CSOPs are not spatiotemporally synchronized and have not yet been considered by current studies. Moreover, GNSS-CSOPs pseudorange measurements noise exhibit differences under the kinematic receiver, and this challenge has not been well dealt with by the existing weighting methods. To address the above-mentioned issues, a novel fusion positioning system based on an iterated extended Kalman filter (IEKF) for GNSS-CSOP is formulated under low observability, and the degree of observability is analyzed. In this system, the influence of spatiotemporal asynchronicity is addressed. Then, a Helmert unit variance estimation (HUVE) algorithm is proposed to obtain the measurement weights in GNSS-CSOPs fusion. Moreover, a double Dog-leg incremental estimation (DDIE) algorithm is proposed to enhance convergence when solving under low observability. In field tests, results show that the positioning accuracy of the proposed system can reach approximately 6.5 m under low observability. Compared with other state-of-the-art studies, such as weighting with power-of-signal, quasi-Newton (QN), Levenberg-Marquardt (LM), and Dog-leg methods, the performance of the proposed system has been improved by 62.4%, 82.4%, 72.9%, and 64.7%, respectively. This study presents a novel fusion positioning strategy for the GNSS-CSOP in low-observability environments.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":null,"pages":null},"PeriodicalIF":5.6,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142595875","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":"Optimization of Space Charge Electro-Optical Detection System in Power Electronic Insulation: Toward High Sensitivity","authors":"Tianrun Qi;Hanwen Ren;Haoyu Gao;Qingmin Li;Xinjun He;Yidan Ma;Yiqun Ma;Tao Xiao","doi":"10.1109/TIM.2024.3485397","DOIUrl":"https://doi.org/10.1109/TIM.2024.3485397","url":null,"abstract":"For high-resolution measurements of solid insulation space charge in complex stress environments, an optical measurement method of space charge based on the electro-optical effect is proposed. In this article, the whole process signal conversion mechanism and transmission model of the space charge electro-optical detection system are constructed. Furthermore, the key factors affecting measurement sensitivity are analyzed for the core electro-optical detection system. The results show that the measurement sensitivity shows a positive and negative correlation with the sensor electro-optical coefficient and sensor thickness, respectively. Simultaneously, it shows a decreasing trend in measurement sensitivity as the angle of incidence deviates from the Brewster angle of the electro-optical sensor. The optimal configuration of the measurement sensitivity can be realized by adjusting the incidence angle of the detection light, the electro-optical coefficient, and the thickness of the sensor. Expanding the measuring bandwidth further decreases the signal distortion. The detector bandwidth is the primary parameter that limits the behavior of the detecting system. Ultimately, a refined design for the electro-optical detecting system is suggested for sensitivity improvement. The significant sensitivity is obtained by incorporating the optical sampling technique. The experimental results demonstrate that the established detection system is capable of accurately measuring the picosecond pulsed electric field. The enhanced electro-optical detection system achieves measurement sensitivity at the \u0000<inline-formula> <tex-math>$mu $ </tex-math></inline-formula>\u0000V level and bandwidth in the THz range. The space charge measurements with nm resolution can be realized through the proposed detection system.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":null,"pages":null},"PeriodicalIF":5.6,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142587603","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}