Search Results (97)

When inverter switching frequency is constrained, PMSM drives operating at relatively high fundamental electrical frequencies may exhibit a low switching-to-fundamental-frequency ratio (low carrier ratio), resulting in fewer voltage vector updates per fundamental cycle and increased current ripple and low-order harmonics. To address the difficulty of synchronizing offline selective harmonic elimination PWM (SHEPWM) switching angles with a field-oriented control (FOC) current loop, this paper proposes an angle-synchronous FOC-SHEPWM implementation based on electrical angle partitioning. A quarter-wave-symmetric SHEPWM model is first established, and the switching angles are solved using Newton iteration and homotopy continuation. A Halton low-discrepancy initial value pool is then constructed. Combined with cumulative interval mapping, admissibility screening, and continuity assessment, it yields switching angle trajectories suitable for closed-loop look-up table implementation and extends the high-modulation-index range toward six-step operation. For online implementation, the ePWM period is updated according to the electrical angular speed, and the offline angles are mapped to intra-partition compare values. Counter-zero sampling, delay angle compensation, dynamic period correction, and fundamental current extraction are integrated to realize synchronized closed-loop pulse generation. The simulation and experimental results demonstrate the stable operation of the SHEPWM-N3, SHEPWM-N5, and SHEPWM-N7 patterns. At equal numbers of switching events, the proposed patterns exhibit lower low-order harmonic content and current total harmonic distortion than ASVPWM, while the fundamental current extraction and dynamic period correction methods reduce d-q-axis current ripple and partition synchronization error, respectively. Full article

No-reference image quality assessment (NR-IQA) metrics are widely used to evaluate and train image and video processing algorithms, but the growing reliance on deep neural networks makes these metrics vulnerable to adversarial attacks. Among such attacks, transferable black-box methods are particularly relevant in real-world scenarios where the attacker cannot interact with the target metric to generate adversarial examples and must rely on a substitute white-box model. In this paper, we examine adversarial transferability across modern NR-IQA metrics and introduce two complementary techniques, Consensus-Guided Distillation (CGD) and Local Perceptual Attack (LPA), which together form a two-stage transferable attack pipeline. CGD distills an ensemble of NR-IQA metrics into a single white-box substitute, excluding training samples for which the teacher metrics show high disagreement, thereby improving transferability and reducing the computational cost of both training and attack generation. LPA optimizes adversarial examples over random multi-scale image partitions, encouraging perturbations to exploit local distortion-sensitive features shared across NR-IQA metrics. Experiments using 15 NR-IQA metrics and four datasets show that distillation substantially improves black-box transferability, while LPA consistently outperforms existing transferable attacks against NR-IQA metrics, both in increasing predicted quality scores and in reducing correlations with subjective quality scores. We also find that attack transferability depends strongly on the type of distortions present in the image, indicating that adversarial perturbations can conceal certain visual degradations from NR-IQA metrics. Full article

Spectral remote sensing struggles in hyper-arid environments because desert substrates share overlapping optical signatures. We evaluate AlphaEarth foundation-model embeddings—64-dimensional annual representations fused from Sentinel-1, Sentinel-2, Landsat, and LiDAR—for land-cover classification and change detection across Saudi Arabia (2017–2024). From an equal-allocation draw of ESA WorldCover 2021 strata, 25,241 labelled samples entered cross-validation; seven classes reached the requested 3000 and the two rarest returned their full national extent at the sampling scale. A Random Forest reproduces the WorldCover labels at a spatially blocked cross-validated overall accuracy of 0.815 ± 0.007, averaged over twenty independent assignments of the spatial blocks to folds, compared with 0.859 ± 0.005 under standard random cross-validation; this figure measures agreement with WorldCover rather than accuracy against independent ground truth, and is not directly comparable with WorldCover’s own globally validated accuracy; the 4.4 percentage-point gap quantifies spatial leakage and is reported transparently. Design weighting following Olofsson et al. corrects the distortion introduced by equal per-class allocation for 2021; because the reference labels are not independent of the training labels, the resulting fractions are reported as model-predicted national composition rather than accuracy-adjusted area estimates. UMAP visualisation of the embedding manifold reveals five sub-types within the single WorldCover bare/sparse vegetation class, consistent with geomorphologically distinct desert substrates. An indicative cross-feature benchmark produced an OA 14.1 percentage points higher for the foundation-model representation than for the strongest Sentinel-2 baseline under the same fold partition. Because the conditions were evaluated on non-identical samples, this difference cannot be attributed solely to feature representation. An independent probability sample of 374 points, interpreted on very-high-resolution imagery in two rounds by two analysts and reconciled to 97.2% agreement, gives a design-weighted overall accuracy of 0.868 for the 2021 map. The same interpretation confirms only 49.5% of the WorldCover class assignments at those points, with tree cover, grassland and herbaceous wetland largely reassigned to cropland, bare/sparse vegetation and shrubland; WorldCover and the Random Forest map agree with the independent reference at the same design-weighted rate of 0.868. The classifier therefore reproduces its label source closely, including where that source departs from independent interpretation, which shows directly that agreement with WorldCover and accuracy against land cover are distinct quantities in this landscape. Together, these results establish a methodologically transparent workflow for foundation model-based land-cover monitoring in data-scarce arid environments, with independent validation limited to the 2021 map and no independent annual reference data available across the complete 2017–2024 period. A separate 49-point cropland two-date sub-test found interpreted field-state change in nine of 25 flagged points (36%) versus two of 24 unflagged points (8%; Fisher’s exact p = 0.037); within this class the screening flag had precision = 0.360, recall = 0.818 and F1 = 0.500, but these class-specific metrics do not constitute national validation of the change layer. Full article

Visual security quantification is critical for evaluating perceptual image encryption schemes. Traditional visual security indices often rely on partial feature extraction and fixed-position block matching, which limits their capability to capture localized information leakage. To address these limitations, this paper proposes a novel Multi-Feature Fusion Visual Security Index (MFFVSI). Within this framework, an optimal block-matching strategy is designed to calculate local feature similarity (LFS), enabling the detection of displaced sensitive information leakage. Concomitantly, global feature similarity (GFS) is derived in the Gaussian-weighted frequency domain to compensate for structural distortion induced by block partitioning. LFS and GFS are subsequently integrated via a multi-resolution regression module to output the final visual security score. Experimental results demonstrate that the proposed method exhibits strong consistency with subjective visual perception and high universality across various perceptual encryption schemes, making it suitable for practical visual security evaluation applications. Full article

The widespread use of location-based services (LBSs) has led to the continuous collection of user location and trajectory data, increasing the risk of privacy leakage and creating a persistent tradeoff between privacy protection and data utility. To address this problem in discrete location query scenarios, this paper proposes a single-point location privacy protection method based on Q-R tree retrieval and differential privacy, termed QRDPP. QRDPP combines the adaptive spatial partitioning capability of a Q-tree with the minimum bounding rectangle (MBR)-based indexing capability of an R-tree. It applies an improved geometric privacy budget allocation strategy to leaf nodes and an arithmetic allocation strategy to non-leaf nodes, followed by Laplace perturbation of the corresponding location data and node information. For continuous trajectory query scenarios, this paper proposes a spatiotemporal generalization and differential privacy method, termed STG-DPTP, to address inadequate temporal protection, inappropriate generalization, and trajectory distortion. STG-DPTP performs hierarchical spatiotemporal clustering, separately models temporal and spatial distributions using Gaussian kernel density estimation, dynamically optimizes bandwidth parameters through Bayesian optimization, selects representative candidate subsets using the exponential mechanism, and generates protected trajectories through constrained sampling. Experiments on the GeoLife dataset evaluate the proposed methods in terms of query accuracy, computational efficiency, spatial trajectory similarity, reconstruction error, adversarial uncertainty, and temporal preservation. The results show that QRDPP improves the utility and efficiency of privacy-preserving spatial queries, while STG-DPTP better preserves the spatial distribution, trajectory structure, and temporal characteristics of the original data under the adopted differential privacy framework. Full article

Clustering may uncover latent behavioral structure in Autism Spectrum Disorder (ASD) screening data without using outcome labels, but the resulting partitions depend strongly on data geometry and the adopted similarity measure. Quantum-hybrid clustering offers alternative distance and similarity estimators, yet whether these subroutines improve on classical methods under controlled conditions remains unclear. We conduct a benchmark of k-means, DBSCAN, agglomerative clustering, and spectral clustering against their quantum-hybrid counterparts. All methods are evaluated in a common 13-dimensional representation, with hyperparameters selected exclusively through internal validation indices. The evaluation covers four synthetic geometries and a 13-dimensional PCA representation of an ASD screening dataset, each containing 300 samples and evaluated over 10 seed-defined stochastic runs. Clustering quality is measured using the Silhouette Index (SI), Davies–Bouldin Index (DBI), Calinski–Harabasz Index (CHI), Adjusted Mutual Information (AMI), and Adjusted Rand Index (ARI). Under this validation protocol, Q-means exactly recovers the Gaussian clusters and improves label agreement on anisotropic data, but it does not outperform classical k-means on the ASD screening data. Q-spectral significantly reduces DBI on Two Moons, Concentric Rings, and ASD screening data, although these reductions do not consistently translate into higher AMI or ARI. Q-DBSCAN and Q-agglomerative exhibit greater sensitivity to distance distortions and finite-shot noise. Overall, the results reveal geometry-dependent trade-offs rather than uniform quantum-hybrid superiority. We relate these findings to theoretical complexity, measurement noise, state-preparation costs, and eigensolver bottlenecks in near-term implementations. Full article

The widespread adoption of Location-Based Services (LBSs) has significantly enhanced daily convenience, yet the upload and storage of user trajectory data pose substantial privacy leakage risks. Addressing the limitations of existing privacy protection methods in achieving personalized adaptation and balancing privacy with utility, this paper proposes a personalized privacy protection strategy for location trajectories based on weighted Kullback–Leibler (KL) divergence. The approach first employs a Markov transition matrix to model user movement patterns, utilizes quadtree-based dynamic grid partitioning for adaptive encoding of the state space, and introduces sensitivity scores weighted by dwell duration and visit frequency to identify critical privacy-sensitive points. It then develops an exponential decay perturbation mechanism combining regularization parameters and distortion thresholds to preserve trajectory spatial usability while protecting sensitive transitions. By quantifying privacy leakage through weighted KL divergence and measuring data utility via distortion metrics, a linearly weighted composite index is constructed, enabling personalized parameter optimization via grid search. Experimental results on the real-world Geolife dataset demonstrate that compared to three differential privacy baselines, this method reduces privacy leakage (measured by weighted KL divergence), improves POI Recall rates, and decreases average geographic errors. Paired t-tests confirm that all improvements are statistically significant (p < 0.001) with large effect sizes, validating its effectiveness and superiority in balancing privacy protection and data usability. Full article

Screen content images (SCIs) exhibit complex structural heterogeneity, rendering traditional statistics-based natural scene image quality assessment (NR-IQA) metrics ineffective. Although deep learning models achieve high prediction accuracy, their prohibitive computational demands preclude deployment in latency-sensitive industrial scenarios. While existing handcrafted lightweight SCI-IQA metrics reduce computational overhead, most rely on unsegmented global feature pooling or holistic edge statistics (e.g., edge histograms or Fisher vector coding), thereby diluting locally critical text-edge distortions in vast homogeneous backgrounds. To address this limitation, we propose an ultra-lightweight, deep-learning-free NR-IQA framework centered on human visual masking. Unlike existing lightweight methods, our approach explicitly employs dual-scale Canny edge operators to partition SCIs into edge-sensitive and flat background regions. Guided by this visual prior, structural degradations and micro-compression textures are extracted region-wise using Sobel gradients and uniform local binary patterns (LBPs) and aggregated with global Commission Internationale de I’Eclairage L*a*b*(CIELAB) color statistics into a compact 60-dimensional descriptor. A grid-search-optimized Support Vector Regression (SVR) maps these features to subjective quality scores. Extensive cross-validation on the SIQAD and SCID datasets demonstrates that our metric outperforms existing handcrafted lightweight SCI metrics and traditional NSS models, while achieving accuracy competitive with representative full-reference metrics. Consuming only 79.3 ms per image on a standard CPU, it offers a practical accuracy–efficiency trade-off for resource-constrained periodic quality monitoring. Full article

Cross-view geo-localization (CVGL) is a critical technology used in unmanned aerial vehicles (UAVs) and widely applied in navigation and target localization tasks. However, owing to the extreme perspective disparity between UAV oblique views and satellite vertical views, CVGL still involves significant challenges, including geometric distortion caused by viewpoint differences, drastic appearance inconsistencies, and the difficulty in bridging semantic gaps between heterogeneous data. To address these issues, we propose a novel CVGL method named dynamic-feature collaborative optimization and semantic-alignment network (DFSA), designed to extract robust feature representations and achieve fine-grained alignment. Specifically, the DFSA employs a residual-based vision transformer as the backbone to capture global context while alleviating the training instability and feature collapse often associated with standard transformers. To bridge the semantic gap between global and local features, we design a feature optimization module comprising a local feature enhancer and a global feature aggregator. This module establishes a closed-loop collaborative system that facilitates top-down semantic guidance and bottom-up detail feedback. Furthermore, we introduce a semantic segmentation and alignment module that adaptively partitions images into semantic regions based on feature response distributions, shifting the matching granularity from the global level to the semantic region level to effectively overcome feature mismatches caused by positional offsets and scale variations. Extensive experiments conducted on the University-1652 and SUES-200 datasets demonstrate the superior image retrieval performance of the proposed DFSA. Specifically, DFSA achieves a Recall@1 of 94.87% and an Average Precision (AP) of 95.32% on the University-1652 dataset and maintains highly competitive Recall@1 performances between 96.83% and 99.25% across various altitudes on the SUES-200 dataset. These results validate the model’s effectiveness in handling extreme viewpoint changes for UAV-based cross-view image retrieval tasks. Full article

Aircraft fuselage corrosion inspection requires both image-domain recognition and metric physical-surface localization for maintenance execution. This study develops a camera-calibrated vision measurement framework that combines PWDE-YOLOv8n-based corrosion perception, original-image coordinate restoration, lens-distortion compensation, ray-based surface mapping, and detector-center bias correction. The perception dataset comprised 2143 images and 5941 corrosion annotations and was partitioned at the physical-specimen, acquisition-session, or source-group level into 1500 training images, 429 validation images, and 214 independent detector-test images. Detailed physical localization was evaluated on a six-image metrology cohort acquired in six sessions, containing 21 corrosion boxes and 84 axial coordinates. A six-fold leave-one-image-out procedure was adopted; in each fold, the center-shift parameters were estimated from the other five images and applied unchanged to the held-out image. The proposed method achieved a mean absolute axial error of 0.641 mm (95% image-cluster bootstrap confidence interval: 0.571–0.708 mm), an RMSE of 0.809 mm, a maximum error of 3.262 mm, and a projected physical-plane bounding-box IoU of 87.12%. The expanded uncertainty of the manually established reference coordinates was 0.374 mm at k = 2, and Monte Carlo propagation produced a mean absolute error of 0.656 mm with a 95% interval of 0.618–0.693 mm. The proposed method reduced the MAE by 97.91% relative to local pixel-to-millimeter scaling and by 70.58% relative to conventional calibrated camera mapping, while producing accuracy comparable to planar homography mapping. Within ρ ≥ 1500 mm, W ≤ 150 mm, and θ ≤ 20°, the estimated curvature-induced additional axial error did not exceed 0.683 mm. A separate ten-image deployment evaluation produced a mean axial error of 2.448 mm and an average processing time of 53.35 ms/image. Full article

Optical remote sensing images acquired in rugged mountains are affected by reflectance distortion caused by both topography and shadows. Most topographic correction (TC) methods normalize sunlit slopes but can become unstable in self shadow and cast shadow, where little or no direct solar radiation reaches the surface. This study builds on the topographic shadow effect correction (TSEC) model and develops TSEC+TC, a partitioned framework for horizontal equivalent normalization. Using a shadow mask extended to penumbra, the framework integrates TSEC for shadowed pixels with conventional TC for sunlit pixels. Both branches target horizontal equivalent reflectance, enabling simultaneous correction of topographic and shadow effects across the scene. We implemented TSEC+TC with path length correction (PLC) and SCS with C (SCSC) models and evaluated it using ten multi-temporal Landsat 8 OLI scenes under different illumination conditions. The results showed that TSEC+TC reduced terrain-related brightness variation and improved land cover classification in the auxiliary comparison relative to uncorrected and TC-only results. For TSEC+SCSC, the R² values between corrected reflectance and cosi were below 0.025 for both Red and SWIR1 bands, and the coefficient of variation of reflectance across aspects was consistently lower than the corresponding values for SE and SCSC, with a maximum of 36.00%. Shadow area analyses indicated that TSEC+TC compensated reflectance distortion in self shadow and cast shadow areas, reduced TC-induced outliers, and better preserved spectral patterns than TC-only correction. Tests using Sentinel-2 MSI and GF-1 WFV imagery provided preliminary evidence of applicability to other sensors. Accounting for the topographic shadow effect improved TC performance in complex mountainous areas. Full article

Underwater object detection is a key technique for marine research. Traditional red green blue (RGB) imaging suffers from short detection ranges and serious color distortion in complex underwater environments. While hyperspectral images contain detailed spectral information, their high dimensionality increases computational costs and noise vulnerability, and high-quality underwater samples are difficult to obtain. A continuity and label-constrained band selection (CLBS) method is presented to address three common defects of existing underwater band selection techniques: target–background confusion, underutilization of spatial features and poor spectral continuity. Three metrics, namely target–background contrast (TBC), target region entropy (Entropy) and spectral–spatial contrast (SSC), are designed for band evaluation. Geometric mean fusion is adopted to suppress extreme values, and spectral continuity constraints are applied to determine the optimal band number. Meanwhile, a multi-category underwater hyperspectral dataset is constructed. Quantitative experiments are carried out on two fully annotated classes (metal and plastic). CLBS reduces the 300 original bands to 209, delivering a 30.3% dimensionality reduction with well-preserved spectral continuity. On a fixed train-validation partition, the 3DCNN+2DCNN model using CLBS-selected bands reaches an F1-score of 94.29%, a Precision of 100.00% and a Recall of 89.19%. Ten repeated tests with random seeds yield averaged results of 94.96 ± 1.13%, 98.44 ± 1.65% and 91.82 ± 3.31% for F1-score, Precision and Recall, respectively, demonstrating reliable performance. Comparative results show that CLBS outperforms conventional feature extraction and various supervised/unsupervised band selection methods. It achieves an excellent trade-off between dimensionality reduction and spectral feature preservation for underwater hyperspectral image processing. Full article

Orthogonal frequency division multiplexing (OFDM) is widely used in broadband wireless communication systems because of its high spectral efficiency, flexible subcarrier allocation, and robustness against frequency-selective fading. However, the high peak-to-average power ratio (PAPR) of OFDM signals reduces the power efficiency of high-power amplifiers and may introduce nonlinear distortion. To address this problem, this paper proposes a multiplier-free Dual-Hadamard PAPR-reduction architecture. The term multiplier-free refers to the additional precoding and phase-weighting operations introduced by the proposed PAPR reduction module, excluding the common OFDM FFT/IFFT operations. Specifically, the proposed architecture combines fast Walsh–Hadamard transform (FWHT) precoding, interleaved partitioning (ILP), and Hadamard-based partial transmit sequence (H-PTS) weighting. Since the FWHT matrix and the H-PTS phase codebook contain only binary signs, the main additional operations are implemented by additions, subtractions, and sign inversions. Simulation results demonstrate that the proposed method reduces the PAPR relative to conventional OFDM without degrading the system’s bit error rate (BER) performance in an additive white Gaussian noise (AWGN) channel when the side information is recovered correctly. Full article

A periodic rotational-scanning panoramic imaging system (PRS imaging system) can acquire large-scale, continuous, and high-resolution panoramic images through rotational scanning. However, non-ideal camera motion during exposure introduces spatially varying motion blur, which degrades image quality and affects subsequent visual interpretation. To address this problem in a self-developed PRS device, this paper proposes an adaptive image deblurring framework based on inertial measurement unit (IMU) assistance and the motion-based motion blur metric (MMBM). First, IMU data collected during exposure are used to calculate the MMBM, which represents the motion blur degree of the current image. The metric is then used to adaptively determine the iteration number of the Richardson–Lucy (R-L) deconvolution algorithm, avoiding unnecessary restoration and reducing artifacts caused by over-restoration. Second, the point spread function (PSF) size is adaptively determined from the camera motion trajectory, and the corresponding PSF is constructed to match different jitter intensities. Finally, an adaptive image partitioning strategy is introduced to handle spatially non-uniform blur caused by camera rotation. Experiments on real images collected by the self-developed dual-spectrum PRS imaging system show that the proposed method achieves stable restoration performance, preserves image naturalness, suppresses unnatural distortions, and reduces computational cost. Full article

This study investigates residual stress development in double-lap T-joints fabricated from medium- and heavy-gauge steel plates. A three-dimensional thermo-mechanically coupled finite element model was developed in Abaqus and validated against blind-hole drilling measurements. Four distinct welding sequence schemes were systematically implemented to quantify their influence on the spatial distribution, peak magnitudes, and evolution trajectories of individual residual stress components (σ_x, σ_γ, σ_z). Results demonstrate that the inherent structural rigidity of medium-to-thick plate assemblies strongly constrains global distortion but does not eliminate sensitivity to sequencing at the local stress level. Although equivalent residual stress peaks remain largely insensitive to welding sequence, the distributions of principal stress components exhibit pronounced sequence-dependent heterogeneity. Specifically, single-side continuous unidirectional welding leverages interpass residual heat accumulation to suppress longitudinal tensile stress, achieving a peak value of 449.9 MPa, the lowest among all configurations. In contrast, double-sided alternating reverse welding promotes thermal dispersion across the joint, thereby reducing both transverse tensile stress magnitude and stress concentration in the distal heat-affected zone. Collectively, these findings establish that optimizing welding sequences for double-lap T-joints in medium-to-heavy plates centers not on minimizing global equivalent stress, but on deliberately tailoring the spatial partitioning and balance of individual stress components, a principle that directly informs robust, performance-driven weld path selection in structural fabrication. Full article