Search Results (201)

Degraded tax forms and other legal–administrative documents require restoration methods that impose strict constraints on content fidelity. This paper presents a structure-aware restoration framework for degraded tax documents that combines geometric normalization, canonical template guidance, and supervised image restoration within a common alignment space. Documents are first mapped to a canonical layout through homography estimation based on Scale-Invariant Feature Transform (SIFT) and Random Sample Consensus (RANSAC), using the canonical visual template as the reference. Restoration is then performed using template priors and masked constraints designed to preserve the fixed document structure while recovering variable content. Under a fixed-budget comparative protocol, U-Net with template priors achieved the highest visual and structural quality, whereas the proposed hybrid model obtained the best functional OCR performance on synthetic data, reaching a Character Error Rate (CER) of 0.1091 and a Word Error Rate (WER) of 0.3495. As a complementary evaluation on 37 real-world documents with human-generated textual ground truth, restoration increased full-page word coverage across all three OCR engines evaluated, yielding absolute improvements ranging from 4.15 to 18.05 percentage points. Full article

Pixel-level annotation of remote sensing imagery is costly, motivating weakly supervised semantic segmentation (WSSS) using only image-level labels. However, class activation maps (CAMs) often highlight only discriminative sub-regions and fail to separate adjacent land-cover regions, particularly in remote sensing scenes characterized by densely co-occurring land-cover classes and substantial variations in object scale. To address these limitations, we propose a three-stage framework that integrates complementary priors from Contrastive Language–Image Pre-training (CLIP), Self-Distillation with No Labels version 2 (DINOv2), and the Segment Anything Model (SAM). First, a lightweight CLIP adapter aligns vision–language priors with remote sensing imagery, while sigmoid-based multi-label decoupled distillation replaces class-competitive distillation with independent class-wise supervision, producing more complete CAMs. Second, DINOv2-guided feature clustering decomposes large merged regions before SAM prompt generation, while Spatial–Semantic Constraints are used to construct confidence-guided point-and-box prompts and reject excessively expanded or semantically inconsistent masks, thereby generating reliable pseudo-labels. Finally, a compact segmentation network is initialized with the weights learned in Stage 1 and retrained using the refined pseudo-labels generated in Stage 2, eliminating the need for foundation models during inference. Experiments on the Potsdam, LoveDA, and DeepGlobe datasets show that the proposed method achieves mean intersection over union (mIoU) scores of 53.16%, 52.66%, and 62.98%, respectively, outperforming state-of-the-art WSSS baselines by 6.55, 1.16, and 1.27 percentage points, respectively. These results demonstrate the effectiveness and generalizability of the proposed framework across diverse remote sensing scenarios under image-level supervision. Full article

Towed hydrophone arrays are widely employed for underwater target detection and direction-of-arrival (DOA) estimation. However, array shape distortion induced by ocean currents, internal waves, and platform maneuvers severely degrades beamforming performance and DOA estimation accuracy. In this paper, a novel Bayesian fusion framework is proposed to achieve robust array shape estimation. Specifically, based on the time-delay estimates derived from the phase differences of line-spectrum components in a pre-processing step, the array geometry is first reconstructed via a piecewise straight-line fitting method. Concurrently, an existing hidden Markov model (HMM)-based method is adopted to estimate the inter-segment deviation angles, in which the smoothness of the array shape is enforced through the state-transition probabilities. The proposed framework then treats these two preliminary estimates as observations from distinct sources and incorporates a smoothness prior within a maximum a posteriori (MAP) formulation that admits a non-iterative closed-form solution to enforce physical continuity constraints on the array geometry. By fusing these complementary estimates, the proposed method simultaneously preserves local sensitivity to fine-scale bends and maintains global consistency of the array shape. Both simulation and lake-trial experiments validate the effectiveness of the proposed method, reducing the array shape estimation error by more than 30% relative to representative existing methods. Moreover, by relying solely on the received acoustic data, the method lowers the dependence on auxiliary sensors and the associated system cost. Full article

To address the inherent limitations of the traditional Aquila Optimizer (AO)—specifically slow convergence, susceptibility to local optima, and limited high-dimensional adaptability—this study proposes a multi-strategy Improved Aquila Optimizer algorithm. Key enhancements include the integration of a logarithmically decaying tangent flight factor to optimize high-dimensional solution distributions, and a dual-layer t-distribution adaptive perturbation model to dynamically regulate search density. Additionally, to solve path-planning problems under strict constraints, we incorporate a prior feasible region initialization, a continuous-to-discrete mapping correction, and a local fine-search mechanism for trajectory smoothing. The proposed Improved Aquila Optimizer algorithm is systematically evaluated against the original AO and six popular algorithms (PSO, SSA, GWO, DBO, DE, and GA) across 23 benchmark functions, the CEC2017 suite, and multi-scale grid maps. The results demonstrate that the Improved Aquila Optimizer algorithm achieves an order-of-magnitude improvement in convergence reliability. By prioritizing absolute search stability and robustness in high-dimensional tasks, the proposed algorithm attains an optimal balance between convergence quality and practical engineering efficiency, proving exceptionally effective in complex path-planning scenarios. Full article

Reviews of artificial intelligence (AI) for mobile robots usually cover one competence—perception, SLAM, path planning, control, or reinforcement learning—and rarely show how these combine into a working system. We take the opposite view and treat autonomy as one pipeline: sensing, perception, localization and mapping, prediction, planning, visual servoing and control, high-level decision-making, and continual learning. We survey how AI has reshaped each stage for industrial and service robots across Industry 4.0, 5.0, and the emerging Industry 6.0. Using a structured, PRISMA-informed protocol with explicit search strings, inclusion criteria, and cross-embodiment transfer rules, we screen the literature, analyze a corpus drawn mainly from the last five years, and position it against prior surveys with a coverage matrix that exposes their single-block focus. Four findings stand out. Perception and localization approach engineering maturity through multimodal fusion and foundation vision models. Planning and control stay effective but computationally demanding. Decision-making, now driven by large language and vision–language–action models, is powerful yet unverifiable and fails under safety constraints. Lifelong learning is almost absent from deployed systems. The decisive weaknesses sit at the interfaces: at the perception–planning, planning–control, and control–decision handoffs the sim-to-real gap, limited on-robot compute, and scarce industrial data compound. We compare AI families by technology readiness, catalog datasets and benchmarks, examine the safety-certification barrier, and consolidate cross-cutting gaps. We close with a staged roadmap toward Industry 6.0 and a next-generation architecture coupling a foundation perception backbone, a world model and digital twin, a continual-learning memory, and a reasoning core wrapped by a safety monitor. The aim is to move from cataloging algorithms to engineering integrated autonomy. Full article

High-precision drogue localization during terminal guidance is critical to close-range autonomous unmanned aerial vehicle (UAV) docking and hinges on infrared–depth (IR–D) multimodal detection. Yet, deploying such detection on airborne edge computing platforms faces severe challenges due to modal heterogeneity, feature redundancy, and real-time constraints. A lightweight IR–D fusion detection network, termed AWIE-CGAF, is proposed for airborne edge deployment, which integrates frequency-domain, physics-prior-driven input enhancement with decoupled gated attention-based adaptive feature fusion to achieve efficient multimodal detection. A training-free Adaptive Wavelet Image Enhancement (AWIE) module is designed to differentially modulate image structures and details in the frequency domain, improving the signal-to-noise ratio and feature discriminability. Concurrently, a Cross-Gated Attention Fusion (CGAF) module employs decoupled cross-modal attention with independent gating, preserving modality-specific features while dynamically selecting complementary information, mitigating redundancy and feature contamination. Experiments on the self-constructed Drogue Infrared–Depth (DIRD) dataset showed that AWIE-CGAF achieved 89.5% mAP@0.5 and 58.2% mAP@0.5:0.95 with 13.5 M parameters, while maintaining real-time inference at 51.7 FPS on a Jetson AGX Orin edge platform. Among the evaluated methods, the proposed framework achieved the highest detection accuracy while retaining real-time edge inference capability. These results support the feasibility of AWIE-CGAF for resource-constrained IR–D drogue perception. Full article

Under complex environmental disturbances and visual measurement noise, globally coherent components, damage-induced local anomalies, and random disturbances are coupled in full-field vertical displacement responses. This coupling limits conventional low-rank–sparse decomposition because of its lack of mechanics-based constraints and can obscure weak damage-induced anomalies. To address this issue, this study proposes a physics-regularized low-rank–sparse damage identification method incorporating a physics prior derived from curvature-strain-energy perturbation. The method first extracts deflection curvature from the full-field displacement responses of the healthy and damaged states. A normalized physical evidence field is then constructed from the curvature-energy difference through Gaussian spatial regularization and mapped into spatially varying sparsity weights to modulate anomaly separation. Subsequently, the Physics-Regularized Differential Damage Index (PRDDI) is constructed from the difference in physics-regularized sparse anomaly intensity between the two states for damage localization and severity-sensitive characterization. The proposed method is primarily intended for beam-like structures satisfying the small-deformation bending assumption. For more complex structures, such as continuous beams, frames, plates, and shells, the corresponding mechanics-based physical evidence and spatial neighborhood relationships can be extended according to their load-transfer mechanisms and spatial geometries. Experimental and numerical results show that the peak-to-background ratio of the physics-regularized sparse anomaly field reaches approximately 2.77 times that of conventional robust principal component analysis (RPCA), while the background level is reduced by approximately 60%, and spurious peaks in non-damaged regions are markedly suppressed. For local stiffness reductions of 5–30%, the PRDDI localization error remains within 0–1 spatial measurement points. Both the peak value and local integrated area within the damaged region increase consistently with the degree of stiffness reduction, with coefficients of determination $R^2$ exceeding 0.99 and Spearman rank correlation coefficients of 1.00. For representative dual-damage cases, the proposed method maintains good dual-peak resolution. Under 10 dB noise, the complete dual-damage detection rate is approximately 87%, while the missed-detection rate for weak damage is approximately 10%. The physics prior derived from curvature-strain-energy perturbation improves consistency with structural mechanics, spatial separability, and the identification reliability of local damage anomaly extraction under complex measurement conditions. By exploiting spatially continuous, vision-based full-field displacement measurements, the proposed method can identify local damage regions in bridges and characterize variations in damage severity, providing a basis for subsequent detailed inspection and condition assessment. Full article

Hydropower units often operate under complex conditions caused by water head change, guide-vane regulation, and load adjustment. These condition changes make it difficult to identify gradual performance degradation from monitoring signals alone. To solve this problem, this paper proposes a condition-aware performance health index construction and multi-source signal-mapping method for hydropower units. First, active power, guide-vane opening, and water head are used as the main operating variables. After data preprocessing and steady-state screening, water head is used as a prior constraint to divide the hydraulic boundary. FCM clustering is then used in each head layer to obtain different operating regions. Second, a high-quantile performance envelope is built in each operating region. The optimal active power is used as the performance benchmark, and the performance health index HI_perf is constructed by comparing actual power with optimal power. The results show that the proposed method can describe the performance deviation under comparable operating conditions. The smoothed daily HI_perf shows a degradation trend before maintenance and a recovery trend after maintenance. Finally, vibration and shaft-swing signals are mapped to HI_perf to construct the signal-based health index HI_sig. The mapping result shows good consistency between HI_sig and HI_perf, and shaft-swing features show stronger sensitivity than vibration features. The proposed framework focuses on daily-scale degradation trend identification using steady-state operating samples, while transient operating events are excluded from the current analysis. The proposed method provides a useful reference for degradation trend identification and health assessment of hydropower units under complex operating conditions. Full article

Detecting naval vessels at extreme distances using infrared search and track (IRST) systems presents severe physical challenges, notably the complete loss of geometric texture and the non-linear submersion of weak target signals within high-dynamic-range sea clutter. Traditional pure data-driven convolutional neural networks (CNNs) rely heavily on visual appearances and suffer from critical feature blind spots under such extreme physical degradation. To overcome this, we propose a Physics-Aware Deep Coupling Network that shifts the detection paradigm from appearance-based feature extraction to physics-guided attribute recognition. Our method deconstructs the degraded infrared signal into three complementary physical domains: an adaptive radiation energy mapping, corresponding to the energy domain, to rescue weak targets; a bio-inspired spatial saliency filtering mechanism, corresponding to the frequency domain, to maximize the signal-to-clutter ratio; and a PSF-coherent gradient topology framework, corresponding to the gradient domain, to discriminate genuine point targets from chaotic sun glints and island edges. These processed priors, alongside the raw image, are integrated into a 4-channel tensor and fused via a Cross-Domain Attention Module, ensuring deep network coupling. To evaluate this architecture, we conduct extensive experiments on the real-world Maritime-SIRST dataset. Since the original dataset provides only pixel-level segmentation masks, we generate axis-aligned bounding-box detection labels from these masks and retrain both the proposed method and a suite of state-of-the-art YOLO detectors under a unified detection paradigm. Extensive benchmarking demonstrates that our physics-aware methodology consistently outperforms these detectors, achieving a mAP50 of 0.923 and an F1 score of 89.92%, thus providing a highly interpretable and robust solution for maritime domain awareness under extreme physical constraints. Full article

Fertilised cropland systems are the dominant source of reactive nitrogen losses in the global environment, contributing most anthropogenic nitrous oxide emissions and driving agricultural nitrogen surplus to nearly three times the safe planetary boundary for nitrogen. This review advances a nitrogen-centred analytical framework that positions cropland nitrogen management as a greenhouse gas mitigation imperative, a planetary boundary challenge, and a cross-cutting lever for integrated Sustainable Development Goal progress. Unlike prior reviews that treat nitrous oxide emissions, nitrogen use efficiency, or Sustainable Development Goal linkages as separate concerns, this framework demonstrates that reducing reactive nitrogen losses from fertilised cropland is the single intervention space most likely to generate co-benefits across food security, clean water, climate action, life on land, and partnership goals, while helping restore the transgressed nitrogen boundary. The nitrogen cascade from fertilised croplands and associated livestock manure generates sequential damages across climate, water quality, human health, and terrestrial biodiversity that are mapped against their goal implications. Mitigation strategies including precision nitrogen management, nitrification inhibitors, biochar amendment, agroforestry, circular manure management, and emerging bio-technical innovations are evaluated against their co-benefits, adoption constraints, and potential to shift losses toward other nitrogen species. A financing gap of over one trillion dollars annually and the absence of binding global nitrogen governance are identified as the primary barriers to integrated action. Spatially explicit nitrogen redistribution and nitrogen use efficiency improvement are presented as design principles capable of reconciling planetary boundary recovery with food security equity across nitrogen surplus and nitrogen-deficit regions. Full article

Traditional Bayesian SAR detection methods suffer poor adaptability to speckle noise, fail to handle severe class imbalance within large-scale multi-target datasets, and incur prohibitive training overheads. To address these drawbacks, this paper develops a lightweight Bayesian detection and recognition framework built upon heavy-tailed Laplacian priors and variational inference. We adopt ResNet-50 as the feature extraction backbone and design a four-stage pipeline: First, a noise-aware Laplacian heavy-tailed prior is proposed to strengthen resistance against speckle outliers. Second, a multi-class variational inference module is constructed to eliminate detection bias induced by uneven sample distribution across target categories. Third, a lightweight uncertainty feedback strategy is introduced to cut computational costs for large-batch training. Evaluated on the MSAR-1.0 dataset, our approach achieves an mAP@0.5 of 94.98% and a macro balanced accuracy ($BA$) of 93.34%. Compared with existing Bayesian detectors, the mAP metric rises by 5.44–6.53%. The model only consumes 4.33 ms per inference frame and completes full training within 1.53 h on a single GPU. Ablation tests validate the independent and combined efficacy of all three core modules. This integrated architecture balances detection precision, classification reliability, and training efficiency, offering a promising prototype for multi-class SAR target interpretation under the evaluated benchmark constraints. Full article

Precise station-keeping of construction barges during offshore operations remains a demanding control problem because the underlying dynamics are highly nonlinear and the disturbance environment is seldom known a priori. This work investigates how a learning-based controller can be embedded into the coordinated winch-control architecture of a specialized engineering vessel to deliver accurate positioning in shallow water. Vessels such as rock-dumping platforms and pipe-laying barges routinely rely on a spread of mooring lines to hold station, and the tensions on these lines are, in current industrial practice, still adjusted manually by the winch operator. The scheme proposed here replaces that manual loop with an adaptive neural feedback law synthesized through backstepping, allowing the unknown portions of the ship model and the exogenous environmental loads to be compensated online without requiring prior identification. The 3DOF control wrench produced by the feedback law is then mapped to the physical line tensions through a constrained allocation that respects the unilateral and breaking-load constraints of the spread. The closed-loop system is shown to be semi-globally uniformly ultimately bounded (SGUUB) in the Lyapunov sense, and its performance is benchmarked against a conventional PD regulator and a nominal model-based design through simulation of a full-scale rock installation barge. When the model-based baseline is given the nominal plant, it attains the cleanest tracking; the proposed neural law achieves comparable steady-state accuracy without requiring prior identification of the hydrodynamic coefficients. A model-free deep reinforcement learning (PPO) controller is additionally benchmarked under irregular (JONSWAP) seas; it attains bounded sub-metre station-keeping without any model knowledge, on par with the PD baseline but less precise than the model-based and adaptive-neural laws. Full article

Existing microseismic prediction methods commonly incorporate domain prior knowledge as ordinary input features, which makes it difficult to fully exploit its structural constraints and physical implications. To address this limitation, this study proposes a geological-structure-prior-guided method (SPG) for predicting the distribution of microseismic energy fields. In the study, a dataset was constructed from 11,081 original microseismic events, yielding 476 day-indexed microseismic energy maps. In SPG, historical microseismic energy maps are used to characterize the recent dynamic evolution of the energy field, whereas geological spatial fields, including coal-seam depth, coal-seam thickness, fault distance, fold distance, and goaf distance, are used to represent static structural priors. A decoder-side prior modulation mechanism is designed to introduce structural constraints into the reconstruction of the target-day energy field. The experimental results show that SPG achieves the best overall continuous-field prediction performance compared with a baseline U-Net and representative spatiotemporal prediction models, including PredRNN-V2, SwinLSTM, and Earthformer. In particular, SPG reduces the mean squared error (MSE) by 11.77% and improves the structural similarity index measure (SSIM) by 2.20% relative to the best-performing comparison model, while also decreasing the average structural gap by 21.29% compared with the baseline U-Net. Furthermore, supplementary high-energy-mask evaluation indicates that SPG maintains competitive capability in identifying high-energy regions. These results indicate that SPG can more effectively reconstruct continuous energy distributions, thereby providing a feasible methodological reference for spatially resolved rockburst early warning and domain-knowledge-informed predictive modeling. Full article

This paper focuses on the indoor robot localization problem in the presence of accumulated errors from odometry and IMU. Considering that the motion of the robot is constrained by the prior map, employing the map as a position reference is beneficial for reducing accumulated errors. Therefore, a particle filtering method with map constraints is proposed to improve indoor robot localization accuracy. First, a LiDAR measurement model is constructed to provide the distance from the robot to map obstacles. Since each particle represents a possible state of the robot, the residual between its virtual measurement and the actual LiDAR measurement is used to update the corresponding particle weight. Additionally, an iterative movement strategy is designed to guide the sampling particles toward the high-probability region for enhancing the effectiveness of the particles. Finally, the experimental results show that the proposed method achieves higher localization accuracy than traditional particle filter methods. Full article

This study focuses on the development of a sensor-informed motion-continuity control method for shared-return electro-hydraulic actuator networks subject to neighboring-branch disturbances. The objective is to reduce the local velocity fluctuations induced by return-line pressure transients while retaining explicit hydraulic and valve constraints. A control-oriented shared-return disturbance model is established to map neighboring-valve action, T-port replenishment, accumulator buffering, common return-line pressure, net driving pressure difference, and local actuator motion. On this basis, sensor-derived motion and pressure states together with neighboring-action prior information are used to reconstruct the objective of a constrained predictive controller according to the disturbance stage. Soft Actor–Critic is restricted to bounded objective-weight inference, whereas the valve command remains generated by locally linearized receding-horizon optimization; bounded mapping, smoothing update, and soft pressure constraints preserve positive weighting matrices and online quadratic programming solvability. Co-simulation, simulation-based ablation and baseline comparisons, timing evaluation, and scaled dual-branch experiments show that the proposed framework improves motion continuity, reduces disturbance-induced pressure-difference excursions, maintains smoother valve execution, and completes each tested online update within the sampling period. These findings support feasibility-preserving sensor-driven objective reconstruction under the investigated shared-return disturbance scenarios. Full article