ieeexplore最新文献

筛选
英文 中文
A Multiscale Attention–Depthwise Network for Microwave Water-Cut Soft Sensing in High-Water-Cut Oil–Water Flow 高含水油水流微波含水软测量的多尺度关注深度网络
IF 4.5 2区 综合性期刊
IEEE Sensors Journal Pub Date : 2026-11-15 Epub Date: 2026-10-13 DOI: 10.1109/JSEN.2026.3741319
Zhiqiang Zhao;Hanqing Chen;Bang Zhou;Zezhong Li;Zhongke Gao
{"title":"A Multiscale Attention–Depthwise Network for Microwave Water-Cut Soft Sensing in High-Water-Cut Oil–Water Flow","authors":"Zhiqiang Zhao;Hanqing Chen;Bang Zhou;Zezhong Li;Zhongke Gao","doi":"10.1109/JSEN.2026.3741319","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3741319","url":null,"abstract":"In high-water-cut oil–water two-phase flow, microwave phase-voltage responses are influenced by dielectric-distribution variations, interfacial fluctuations, and flow-structure changes, exhibiting strong nonlinear and dynamic characteristics that challenge conventional calibration-based water-cut estimation methods. To address this issue, a directional antenna microwave sensor (DAMS) based on a printed half-wavelength dipole is developed to acquire dynamic phase-voltage time-series responses. Based on the acquired phase-voltage sequences, a multiscale attention–depthwise network (MSADNet) is proposed for microwave water-cut soft sensing. The proposed network integrates multiscale temporal feature extraction, attention–depthwise mixing, and cross-gating feature fusion to capture local signal fluctuations and long-range temporal dependencies. Dynamic experiments are conducted for upward oil–water flow in a vertical loop under water cuts of 70%–95% and total flow rates of 0.36–1.20 m<sup>3</sup>/h. Within the investigated operating range, MSADNet achieves a root-mean-square error (RMSE) of <inline-formula> <tex-math>$1.80~pm ~0.12$ </tex-math></inline-formula> percentage points (p.p.), a mean absolute error (MAE) of <inline-formula> <tex-math>$1.14~pm ~0.11$ </tex-math></inline-formula> p.p., and a mean absolute percentage error (MAPE) of 1.43% <inline-formula> <tex-math>$pm ~0.14$ </tex-math></inline-formula>%, outperforming the evaluated conventional and deep-learning-based methods. These results indicate the potential of combining directional antenna microwave sensing with deep temporal representation learning for water-cut soft sensing under dynamic high-water-cut flow conditions.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 22","pages":"34868-34878"},"PeriodicalIF":4.5,"publicationDate":"2026-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149415979","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Noninvasive RPM Estimation Using Spintronic Sensors With Spectral Peak Tracking 基于谱峰跟踪的自旋电子传感器无创转速估计
IF 4.5 2区 综合性期刊
IEEE Sensors Journal Pub Date : 2026-11-15 Epub Date: 2026-10-14 DOI: 10.1109/JSEN.2026.3741702
Ensieh Iranmehr;André Araújo;Ricardo Ferreira
{"title":"Noninvasive RPM Estimation Using Spintronic Sensors With Spectral Peak Tracking","authors":"Ensieh Iranmehr;André Araújo;Ricardo Ferreira","doi":"10.1109/JSEN.2026.3741702","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3741702","url":null,"abstract":"Noninvasive motor rotational speed (RPM) sensing is of increasing interest for electric vehicle (EV) monitoring, predictive maintenance, and intelligent transportation systems. This article presents an RPM estimation framework that combines tunneling magnetoresistance (TMR) sensing with a context-aware spectral peak tracking approach, validated using real EV operating data. Externally mounted TMR sensors capture magnetic field signatures generated by motor operation, enabling speed estimation without mechanical modifications or direct access to vehicle control signals. The acquired signals are processed in the frequency domain, where motor-related spectral components are identified and tracked over time using temporal continuity and spectral context. The proposed approach provides a practical methodology for extracting motor rotational speed from externally measured magnetic field signals under varying operating conditions. Experimental validation using data collected from a real EV platform yielded an <inline-formula> <tex-math>$R^{2}$ </tex-math></inline-formula> of 0.94 and a normalized root-mean-square error (NRMSE) of 6.5% relative to the reference motor RPM. These findings demonstrate the feasibility of TMR-based motor speed estimation and highlight its potential for automotive monitoring applications.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 22","pages":"34346-34356"},"PeriodicalIF":4.5,"publicationDate":"2026-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149415991","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Self-Referenced Dual-EGFET Platform for pH-Compensated Urea Detection 自参考双egfet平台用于ph补偿尿素检测
IF 4.5 2区 综合性期刊
IEEE Sensors Journal Pub Date : 2026-11-15 Epub Date: 2026-10-07 DOI: 10.1109/JSEN.2026.3738521
Ashok Kumar Nanda;C. Venkata Rambabu;Sunkara Teena Mrudula;K. V. Balaramakrishna;Srinivasulu A;D. Tilak Raju;G. Lakshmana;Muralidharan J;Ramesh Babu Yathirajula
{"title":"Self-Referenced Dual-EGFET Platform for pH-Compensated Urea Detection","authors":"Ashok Kumar Nanda;C. Venkata Rambabu;Sunkara Teena Mrudula;K. V. Balaramakrishna;Srinivasulu A;D. Tilak Raju;G. Lakshmana;Muralidharan J;Ramesh Babu Yathirajula","doi":"10.1109/JSEN.2026.3738521","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3738521","url":null,"abstract":"Enzymatic urea sensors based on pH transduction are vulnerable to variation in sample bulk pH. This work reports a self-referenced dual extended-gate field-effect transistor (EGFET) platform comprising a bare tantalum nitride (TaN) channel for independent bulk-pH measurement and a urease-functionalized TaN channel for urea-induced local-pH transduction. Both extended gates were measured sequentially with the same nMOS transducer, thereby avoiding transistor-to-transistor mismatch. The measured bulk pH selects or interpolates the functionalized channel’s pH-specific zero-urea baseline, which is subtracted before concentration estimation. Surface microscopy, Fourier-transform infrared spectroscopy (FTIR), X-ray photoelectron spectroscopy (XPS), and four-point-probe measurements verify the immobilized enzyme layer and the conductive, oxidized TaN sensing interface. The bare channel exhibited a pH sensitivity of <inline-formula> <tex-math>$56.4pm 0.6~text {mV/pH}$ </tex-math></inline-formula> with <inline-formula> <tex-math>$text {R}^{2}=0.997$ </tex-math></inline-formula>. The corrected urea response was linear from 0.05 to 10 mM, was tested up to 20 mM, and yielded <inline-formula> <tex-math>$55.42pm 0.26~text {mV/dec}$ </tex-math></inline-formula>, <inline-formula> <tex-math>$text {R}^{2}=0.996$ </tex-math></inline-formula>, and an estimated limit of detection (LOD) of 0.043 mM. The 90% response time was 55–72 s, hysteresis was <inline-formula> <tex-math>$3.58pm 0.36~text {mV}$ </tex-math></inline-formula>, and 95.8% of the initial response was retained after 21 days. Across six pH–concentration conditions, compensation reduced mean estimation error from 24.6% to <inline-formula> <tex-math>$5.1%~(text {p}lt 0.001)$ </tex-math></inline-formula>. Fixed-pH controls produced maximum bare-channel shifts of 1.3 mV over 10–200 mM NaCl/KCl and 0.9 mV after representative urine-metabolite additions. Preliminary spiked artificial-urine recoveries were 94.2%–106.0%. The results establish baseline-anchored self-referencing while defining the calibrated matrix, temperature, and validation boundaries of the platform.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 22","pages":"33852-33859"},"PeriodicalIF":4.5,"publicationDate":"2026-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149416012","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
UEPD: Underwater Efficient Pipeline Object Detector UEPD:水下高效管道目标检测器
IF 4.5 2区 综合性期刊
IEEE Sensors Journal Pub Date : 2026-11-15 Epub Date: 2026-10-09 DOI: 10.1109/JSEN.2026.3740098
Xue-Han Xu;Qing-Hao Meng;Tao Jing;Ming Zeng
{"title":"UEPD: Underwater Efficient Pipeline Object Detector","authors":"Xue-Han Xu;Qing-Hao Meng;Tao Jing;Ming Zeng","doi":"10.1109/JSEN.2026.3740098","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3740098","url":null,"abstract":"Efficient detection of underwater pipeline targets is crucial for unmanned underwater vehicles (UUVs) to conduct subsea infrastructure inspection. However, underwater optical sensing is highly sensitive to environmental degradations. Although current detectors have reported encouraging results, robust detection remains challenging due to the severe loss of structure-sensitive cues in degraded optical sensor data, especially when real-time inference is required on resource-limited UUV edge platforms. In this article, an underwater efficient pipeline detector (UEPD) is proposed. To enhance degraded structural representations, an edge-guided frequency enhancement module (EFEM) is embedded into the backbone, where wavelet-based frequency decomposition is combined with edge-guided high-frequency modulation and low-frequency conditional reconstruction. To further improve structural discrimination during feature aggregation, a boundary-region guided feature fusion network (BRGFN) is introduced in the neck to explicitly model the complementary effects of region-aware and boundary-aware cues. To validate the performance of the proposed method, a Turbid Water Pipeline Dataset (TWPD) is collected from real sea areas, which includes background images and four types of targets: pipelines, text, pipeline connections, and metal defects. Experimental results demonstrate that UEPD provides a favorable balance between detection accuracy and computational efficiency, indicating its potential for practical deployment in underwater pipeline inspection systems. The code is available at <uri>https://github.com/xxh1016/UEPD</uri>","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 22","pages":"34763-34773"},"PeriodicalIF":4.5,"publicationDate":"2026-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149415847","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Secure and Energy-Efficient Routing in Biomedical Wireless Sensor Networks Using Anomaly Detection and Trust-Based Key Management 基于异常检测和基于信任的密钥管理的生物医学无线传感器网络安全节能路由
IF 4.5 2区 综合性期刊
IEEE Sensors Journal Pub Date : 2026-11-15 Epub Date: 2026-07-10 DOI: 10.1109/JSEN.2026.3709829
Ahmet Berker Memisoglu;Gorkem Memisoglu
{"title":"Secure and Energy-Efficient Routing in Biomedical Wireless Sensor Networks Using Anomaly Detection and Trust-Based Key Management","authors":"Ahmet Berker Memisoglu;Gorkem Memisoglu","doi":"10.1109/JSEN.2026.3709829","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3709829","url":null,"abstract":"Wireless sensor networks (WSNs) are increasingly integral to healthcare automation by enabling real-time patient monitoring and remote diagnostics. However, their deployment in resource-constrained environments exposes them to serious security threats such as false data injection attacks (FDIAs), node compromise, or cryptographic key leakage. This study presents a secure, anomaly-aware, and energy-efficient WSN framework for healthcare applications, using real glucose sensor data collected from five sensor nodes (SNs) over 100 days. The proposed system integrates advanced encryption standard, elliptic curve cryptography (ECC)/Rivest–Shamir–Adleman (RSA)-based authentication, and a trust-based secure routing (TBSR) scheme combined with an intrusion detection system leveraging z-score-based anomaly detection. Dynamic key management techniques are incorporated to reduce key compromise risks over time. Python-based simulations demonstrate high-detection rates, low-cryptographic latency, and balanced energy consumption across different encryption schemes and key sizes. Results also highlight that dynamic keying significantly reduces compromise probability compared to static keying. The visualization of anomaly scores, encryption timings, and energy use confirms the robustness of the proposed framework in preserving data confidentiality, integrity, and availability under attack scenarios. This model serves as a practical foundation for the development of secure next-generation healthcare WSNs.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 22","pages":"35188-35203"},"PeriodicalIF":4.5,"publicationDate":"2026-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149415854","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
TJBe-GS: Rigid Trajectory-Guided Bézier Gaussian Splatting for RGB-D Dynamic Reconstruction RGB-D动态重建的刚体轨迹制导b2013.zier高斯溅射
IF 4.5 2区 综合性期刊
IEEE Sensors Journal Pub Date : 2026-11-15 Epub Date: 2026-09-21 DOI: 10.1109/JSEN.2026.3734149
Wei Wei;Chao Li;Xianghui Pan;Jiayuan Du;Chengju Liu;Qijun Chen
{"title":"TJBe-GS: Rigid Trajectory-Guided Bézier Gaussian Splatting for RGB-D Dynamic Reconstruction","authors":"Wei Wei;Chao Li;Xianghui Pan;Jiayuan Du;Chengju Liu;Qijun Chen","doi":"10.1109/JSEN.2026.3734149","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3734149","url":null,"abstract":"Dynamic scene reconstruction via vision sensors is a crucial basis for accurate spatiotemporal environment perception and geometric modeling. Against this backdrop, accurately modeling the continuous pose variations of dynamic objects over time is crucial for applications such as simultaneous localization and mapping (SLAM), autonomous driving, and robot imitation learning. However, existing 3-D Gaussian splatting (3DGS) methods couple object-level rigid motion with local deformation, which increases motion ambiguity and may cause trajectory drift, dragging artifacts, or blurred local geometry under complex RGB-D dynamics. To address this, we propose TJBe-GS, a novel 4-D reconstruction framework for precise dynamic modeling. Unlike traditional mesh- or anchor-based methods, we formulate the continuous evolution of dynamic objects as a superposition of global rigid displacements and local nonrigid deformations. Specifically, we leverage synchronized RGB-D streams, and instance masks are used to separate dynamic objects from the static background, acquiring precise point-prior trajectories for the dynamic elements by a geometrically guided initialization and filtering module (GIFM). We then establish static and dynamic Gaussian primitives: static primitives are enhanced via temporal accumulation and random downsampling, while each dynamic Gaussian is assigned a learnable Bézier curve to model continuous local deformation. Furthermore, we introduce a motion smoothness constraint to enforce physical plausibility and temporal coherence. Extensive experiments and ablation studies conducted on the TUM, BONN, and custom datasets demonstrate that TJBe-GS reconstructs dynamic objects with superior quality and precise geometric details compared to existing methods. Demos and code are available at <uri>https://github.com/wei17861101975-hash/TJBe-GS</uri>","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 22","pages":"34403-34416"},"PeriodicalIF":4.5,"publicationDate":"2026-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149415859","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
AI-Driven Instrumentation for Dual-Sided Coated Quartz Crystal Microbalance for Detecting Multiple Environmental Parameters 用于检测多种环境参数的双面涂层石英晶体微天平的人工智能驱动仪器
IF 4.5 2区 综合性期刊
IEEE Sensors Journal Pub Date : 2026-11-15 Epub Date: 2026-10-01 DOI: 10.1109/JSEN.2026.3737150
Chin Leon Chung;Wei Yin Lim;Narayanan Ramakrishnan
{"title":"AI-Driven Instrumentation for Dual-Sided Coated Quartz Crystal Microbalance for Detecting Multiple Environmental Parameters","authors":"Chin Leon Chung;Wei Yin Lim;Narayanan Ramakrishnan","doi":"10.1109/JSEN.2026.3737150","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3737150","url":null,"abstract":"Environmental monitoring often requires simultaneous detection of multiple parameters, as single-parameter sensors are insufficient for comprehensive analysis. Multiplex sensors, which are cost-effective and time-efficient, can provide a practical solution. By leveraging machine learning techniques, measurement precision can be further improved through the development of models capable of delivering accurate predictions. In this study, a dual-sided coated quartz crystal microbalance (QCM) sensor was developed for simultaneous monitoring of multiple environmental parameters, including ultraviolet (UV) intensity and relative humidity (RH). Both sides of each electrode of the QCM were coated with graphene oxide (GO) and zinc oxide (ZnO) to enhance detection capabilities, and measurements were conducted using series resonance frequency (<italic>F</i><sub>s</sub>), series resonance impedance (<italic>Z</i><sub>s</sub>), parallel resonance frequency (<italic>F</i><sub>p</sub>), and parallel resonance impedance (<italic>Z</i><sub>p</sub>). Three deep learning architectures—convolutional neural network (CNN), gated recurrent unit (GRU), and long short-term memory (LSTM)—were developed to model the complex QCM data. Among various models tested, the deep learning algorithm GRU outperformed traditional machine learning regression methods, achieving the highest R<sup>2</sup> score of 0.798 and the lowest mean squared error (mse) of 27.433. The results demonstrate the potential of dual-sided coated QCM sensors integrated with advanced machine learning for precise, multiparameter environmental monitoring. This approach could be further extended for gas sensing and Internet of Things (IoT) applications.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 22","pages":"34013-34022"},"PeriodicalIF":4.5,"publicationDate":"2026-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149415866","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Abnormal Sample Identification in Forensic Cartridge-Case Analysis Using a Cascaded DBSCAN--PointNet Model 使用级联DBSCAN- PointNet模型的法医墨盒案例分析中的异常样本识别
IF 4.5 2区 综合性期刊
IEEE Sensors Journal Pub Date : 2026-11-15 Epub Date: 2026-10-08 DOI: 10.1109/JSEN.2026.3739959
Yong Guo;Kejin Chen;Linyu Huang;Qian Ning;Yongsheng Li
{"title":"Abnormal Sample Identification in Forensic Cartridge-Case Analysis Using a Cascaded DBSCAN--PointNet Model","authors":"Yong Guo;Kejin Chen;Linyu Huang;Qian Ning;Yongsheng Li","doi":"10.1109/JSEN.2026.3739959","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3739959","url":null,"abstract":"Line-laser sensors are widely used in forensic trace analysis due to their noncontact and high-precision measurement capabilities. In cartridge-case examination, line-laser scanning is employed to acquire 3-D topography of firing pin impressions on cartridge-case primers. However, the complex geometry of firing pin impressions and the highly reflective metallic surface of cartridge cases often introduce measurement anomalies, which may compromise subsequent forensic analysis. To address this issue, a cascaded anomaly-identification framework integrating density-based spatial clustering of applications with noise (DBSCAN) and an improved PointNet is proposed. DBSCAN is first used to directly identify discontinuous anomalous samples containing scattered points, outliers, and discontinuous point-cloud distributions. The remaining samples are then analyzed by an improved PointNet with a dual-pooling strategy to distinguish continuous anomalous samples from normal samples. Experiments conducted on a self-collected dataset containing 3281 firing-pin-impression point clouds demonstrate that the proposed framework achieves accuracies of 99.26% and 100% on the training and test sets, respectively. The proposed method provides an effective automated solution for measurement-anomaly screening in forensic cartridge-case acquisition workflows and shows potential for extension to other 3-D measurement applications involving highly reflective metallic surfaces.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 22","pages":"34737-34748"},"PeriodicalIF":4.5,"publicationDate":"2026-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149415877","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Portable Multispectral Optical Reader for Paper-Based Phosphate Detection in Saliva 唾液中纸基磷酸盐检测的便携式多光谱光学读取器
IF 4.5 2区 综合性期刊
IEEE Sensors Journal Pub Date : 2026-11-15 Epub Date: 2026-09-22 DOI: 10.1109/JSEN.2026.3734592
Rishabh Tomar;s Rohit;Sachin Kumar;Venencia Albert;Deepak Agrawal;Abhishek Raj
{"title":"Portable Multispectral Optical Reader for Paper-Based Phosphate Detection in Saliva","authors":"Rishabh Tomar;s Rohit;Sachin Kumar;Venencia Albert;Deepak Agrawal;Abhishek Raj","doi":"10.1109/JSEN.2026.3734592","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3734592","url":null,"abstract":"Phosphate quantification remains constrained by the reliance on conventional laboratory spectrophotometers and microplate readers, which require controlled laboratory infrastructure and are less suitable for simple field deployment, limiting the practicality of decentralized or noninvasive screening. This work introduces a novel portable multispectral optical reader that integrates wax-patterned paper-based analytical devices PADs (<inline-formula> <tex-math>$mu $ </tex-math></inline-formula><italic>PADs</i>) with vanadomolybdate yellow colorimetry to achieve robust onsite phosphate detection. Unlike camera-based colorimetry, the proposed system leverages narrow-band spectral measurements from an AS7341 multispectral sensor, combined with controlled white-LED illumination, dual diffuser-assisted uniform lighting, and a 3-D-printed light-shielded optical enclosure, enabling illumination-independent, ratio-metric spectral analysis. Distinct spectral features extracted from 480 to 590 nm channels-particularly the F4/F3 ratio exhibited strong linearity across 50–120 mg/L (<inline-formula> <tex-math>$mathrm{R}^{{2}} approx {0.98}$ </tex-math></inline-formula>). Validation using standard solutions and processed saliva from 38 volunteers demonstrated close agreement with a laboratory microplate reader, with average deviations of 11.34% for human saliva samples. The analytical sensitivity yielded a limit of detection (LOD) of 11.86 mg/L and a limit of quantification (LOQ) of 35.96 mg/L. By combining <inline-formula> <tex-math>$mu $ </tex-math></inline-formula><italic>PAD</i> simplicity with multispectral precision, this work presents an integrated, low-cost, noninvasive phosphate sensing platform using ratio-metric multispectral reflectance, establishing a practical alternative to conventional laboratory instrumentation for onsite and salivary phosphate monitoring.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 22","pages":"34039-34046"},"PeriodicalIF":4.5,"publicationDate":"2026-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11703905","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149415889","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
S2TANet: SAM2-Based Spatiotemporal-Aware Network for Remote Sensing Change Detection S2TANet:基于sam2的遥感变化检测时空感知网络
IF 4.5 2区 综合性期刊
IEEE Sensors Journal Pub Date : 2026-11-15 Epub Date: 2026-10-01 DOI: 10.1109/JSEN.2026.3736968
Yujie Zhang;Jue Wang;Wenchao Liu;Guoqing Wang;Tianyu Wei;Liang Chen
{"title":"S2TANet: SAM2-Based Spatiotemporal-Aware Network for Remote Sensing Change Detection","authors":"Yujie Zhang;Jue Wang;Wenchao Liu;Guoqing Wang;Tianyu Wei;Liang Chen","doi":"10.1109/JSEN.2026.3736968","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3736968","url":null,"abstract":"In recent years, deep learning-based methods for remote sensing change detection (CD) have developed rapidly. However, radiometric differences between bitemporal images caused by seasonal or illumination variations often make changed regions difficult to detect accurately. Motivated by its multiscale segmentation-oriented feature representations, we introduce the frozen segment anything model 2 (SAM2) image encoder into the framework. Based on this encoder, we propose an SAM2-based spatiotemporal-aware (STA) CD network (S<sup>2</sup>TANet). S<sup>2</sup>TANet integrates two components: an STA adapter and a dynamic adaptive scanning change (DASC) decoder. The STA adapter extracts difference features from bitemporal images and enhances them through spatial attention. In a single forward pass, the attention-guided difference features are fed back into the bitemporal branches. This operation enhances feature consistency in unchanged regions while enlarging feature discrepancies in changed regions, thereby mitigating radiometric interference. The DASC decoder is composed of multiple DASC blocks. Each block employs a dynamic adaptive scanning mechanism. Specifically, an offset prediction network (OPN) predicts offsets for scanning reference points from local features. The predicted offsets displace the scanning reference points to locate patches of interest and adapt the scanning strategy. This content-dependent scanning enhances the detection of irregular changes under complex backgrounds. Extensive experiments on SYSU-CD, SECOND-CD, and LEVIR-CD show that S<sup>2</sup>TANet outperforms the strongest competing methods. The corresponding <inline-formula> <tex-math>$F1$ </tex-math></inline-formula>-score gains are 1.87, 2.00, and 0.43 percentage points, respectively. Code is available at <uri>https://github.com/bitzyj/S2TANet</uri>","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 22","pages":"34506-34520"},"PeriodicalIF":4.5,"publicationDate":"2026-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149415900","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
相关产品
×
本文献相关产品
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信
小红书