搜索结果：Vision (Basel, Switzerland)

The global cattle industry is experiencing significant growth, requiring advanced methods for monitoring animal health and welfare to ensure productivity and sustainability. Traditional manual monitoring techniques are labor-intensive and often impractical for large-scale operations. This review provides a comprehensive analysis of existing and emerging computer vision tools applied to the monitoring of cattle health and welfare. By systematically examining studies across major databases, this paper addresses six key research questions focusing on (1) the issues addressed by computer vision technologies, (2) data acquisition systems, (3) implemented techniques and algorithms, (4) performance outcomes, (5) challenges faced, and (6) potential applications for underexplored health and welfare aspects in cattle farming. The findings show that computer vision technologies have significantly progressed in areas such as body condition score detection, lameness detection, weight estimation, estrus detection, monitoring of feeding and drinking behavior, breathing detection, and recognition of general behaviors. Despite the progress, challenges such as variability in environmental conditions, the need for large annotated datasets, and the high cost of advanced imaging equipment persist. The review emphasizes future research opportunities to address these challenges by focusing on disease-specific monitoring. This review aims to provide veterinarians, farmers, and animal health professionals with greater insight into computer vision technologies and to promote their adoption by discussing their practical applications.

Deploying deep vision models on edge hardware requires understanding how architecture and training hyperparameters jointly determine accuracy and inference latency, yet these relationships remain poorly characterized in a systematic, data-driven manner. This paper presents a two-stage statistical framework providing interpretable, closed-form insights into both. In the first stage, we apply distance correlation (dCor) and the maximal information coefficient (MIC) across seven image-classification datasets, revealing that batch size and total layer count are the strongest universal accuracy predictors (mean dCor: 0.228 and 0.174), while learning rate achieves the highest MIC (0.226), reflecting a non-monotonic relationship with accuracy. In the second stage, PySR symbolic regression (representing, to our knowledge, the first application to cross-dataset vision model accuracy prediction) derives compact, interpretable formulas. Dataset-specific models achieve R2 from 0.20 to 0.45; a universal model achieves a mean leave-one-dataset-out R2 of 0.23, remaining strictly positive on all held-out datasets, whereas ordinary linear regression collapses to R2=-0.71. We further derive device-specific inference latency formulas for CPU, GPU, and NPU, outperforming classical baselines by 6.7×-14.8× in R2 and confirming fundamental device heterogeneity. Together, these results offer interpretable surrogate models for screening deep vision architectures under accuracy and latency constraints in edge deployment.

Autonomous Haulage Systems (AHS) have significantly transformed surface mining operations by improving safety, productivity, and operational consistency. Currently, AHS predominantly rely on vehicle-centric perception architectures. Onboard LiDAR, radar, cameras, and Global Navigation Satellite Systems (GNSS) perform sensing, interpretation, and decision-making within individual systems. These processes enable collision avoidance and path tracking. However, they are limited in their ability to consider the broader, dynamic mining environment characterized by dust, terrain degradation, geotechnical instability, heterogeneous traffic, and rapidly evolving operational conditions. This paper presents a systematic review of dynamic vision systems of AHS in surface mining. It critically analyzes the transition from autonomy to interconnected, ecosystem-aware intelligence. The review synthesizes literature from mining automation, robotics, intelligent transportation systems, and multi-agent perception. It assesses sensing technologies, perception algorithms, sensor fusion strategies, and environmental robustness techniques. Attention is focused on the limitations of egocentric perception models in complex surface mining ecosystems. Building on identified gaps, the paper proposes a conceptual framework for Ecosystem-Centric Dynamic Vision (ECDV). Perception is enhanced through integration with fleet communication networks, dispatch systems, digital twins, geotechnical monitoring platforms, and environmental sensing infrastructure. The framework outlines a multi-layer architecture enabling cooperative perception, predictive hazard modeling, and risk-aware decision support at the mine-wide level. The review concludes by outlining a research agenda to transition from vehicle autonomy to ecosystem intelligence in surface mining. It highlights opportunities in cooperative perception, adaptive sensor fusion under degraded visibility, and digital-twin-integrated predictive safety systems.

Microscopic vision-based alignment measurement is a key procedure in micro-/nanoscale positioning, and its measurement repeatability mainly depends on the stability of subpixel edge-center estimation. However, in practical microscopic imaging, defocus and contamination can cause edge broadening and pseudo-gradient peaks, making it difficult for conventional methods to accurately estimate the edge center of alignment marks. To address this problem, this paper proposes an adaptive edge-response modeling method. First, an amplitude function is constructed by combining the gradient peak and the slope of the edge-transition region, enabling adaptive adjustment of the response amplitude and suppressing its coupling with other parameters. On this basis, the proposed model overcomes the limitation that the Sigmoid model is only suitable for single-edge fitting and enables unified modeling of practical multi-edge hybrid bonding marks. It also suppresses the interference caused by edge pseudo-peaks and abrupt gradient variations, thereby improving the accuracy of subpixel fitting and localization. Experimental results show that, compared with conventional methods, the proposed method improves the repeatability of subpixel edge localization under degraded microscopic imaging conditions by approximately 52%, meeting the requirements of high-precision microscopic vision-based alignment.

Background/Objectives: Manual Pap smear analysis for cervical cancer screening is limited by inter-observer variability, time constraints, and restricted expert availability. Although convolutional neural networks (CNNs) have automated cervical cell classification, they remain limited in modeling long-range spatial dependencies and often lack clinical interpretability. Methods: In this study, Vision Transformer (ViT) architectures were systematically optimized to enhance automated cervical cancer screening and improve interpretability. The Herlev dataset (917 images: 242 normal, 675 abnormal) was utilized to optimize ViT-Tiny, a lightweight ViT architecture designed for reduced computational complexity, through a comprehensive evaluation of augmentation strategies, class weighting, and hyperparameters. Results: The optimal configuration achieved a cross-validation accuracy of approximately 95% (94.89% for the best replicated configuration), in which random horizontal flipping and class weighting (0.7 × 1.3) were identified as most effective. Gradient-weighted Class Activation Mapping (Grad-CAM) analysis confirmed that model attention corresponded to clinically relevant morphological features, including nuclei regions, cell boundaries, and chromatin texture, which align with cytopathological criteria. Conclusions: These findings indicate that Vision Transformers can deliver accurate and interpretable decision support for cervical cancer screening by combining competitive classification performance with attention-based transparency relevant to medical AI. Further validation on larger, multi-center datasets remains necessary before clinical deployment.

Medical image analysis is a cornerstone of modern healthcare, yet conventional single-modal deep learning often struggles with the unique physical constraints and structural variability inherent in data acquired from diverse medical sensors. Recently, Vision-Language Models (VLMs) have sparked a paradigm shift by bridging the semantic gap between visual sensor signals and clinical narratives. Following the PRISMA guidelines, 167 representative studies are systematically synthesized in this review to provide a comprehensive roadmap of VLM technological evolution and clinical utility. First, rather than treating VLMs as generic feature extractors, their underlying mechanisms are uniquely distilled into seven core operational principles, which are then explicitly mapped to downstream applications such as few-shot diagnosis, prompt-driven segmentation, and multi-task foundation models. To facilitate intuitive evaluation, a rigorous quantitative cross-comparison of current benchmark architectures is presented. Crucially, this review goes beyond highlighting successes by critically assessing prevalent clinical bottlenecks, including zero-shot segmentation failures, multi-modal hallucinations in diagnosing rare diseases, and the prohibitive computational complexity associated with 3D volumes and gigapixel whole slide images. Finally, a novel, forward-looking framework is proposed: the transition from static "image-text alignment" to dynamic "multi-source sensor-driven intelligence". By addressing both physical sensor constraints and algorithmic limitations, this survey offers actionable insights for developing trustworthy, sensor-aware clinical diagnostic agents.

Automated vision-based sensing for personal protective equipment (PPE) compliance in high-formwork support system (HFSS) construction environments faces three compounding challenges: extreme within-image scale variation, dense scaffold occlusions, and small safety hook targets prone to missed detection. Existing sensing systems address only binary presence detection and cannot assess whether safety harnesses are anchored in compliance with regulatory requirements. This paper proposes YOLO-ILB, a lightweight task-specific object detector built on YOLO11n with three targeted improvements. The C3k2_IDWC module replaces standard convolutions with multi-branch Inception depthwise convolutions, improving multi-scale feature discrimination at reduced computational cost. The SPPF_LSKA module embeds large separable kernel attention into the SPPF aggregation path, strengthening global context awareness to suppress scaffold background interference. A BiFPN neck replaces the original PAN structure, enabling bidirectional cross-scale weighted feature fusion to balance detection of small hooks and large harnesses within a sin gle image. A UAV-based sensing dataset was constructed using a DJI Mini 3 Pro (4032 × 3024 px) across 17 real construction sites under varied altitudes, viewing angles, and illumination conditions, yielding 2700 annotated images across five object categories. YOLO-ILB achieves mAP50 = 0.939 with only 1.923 M parameters and 5.7 G FLOPs at 262.3 FPS, outperforming eight mainstream YOLO baselines while remaining deployable on resource-constrained edge computing nodes. A geometry-based compliance algorithm further classifies three harness anchoring states-correct high anchoring, incorrect low anchoring, and unclipped or excessively distant hook-without additional sensors or annotations, achieving 90.82% overall accuracy on 305 field instances and extending the sensing system from presence detection to regulatory compliance assessment.

Background/Objectives: Impacted third molars may adversely affect adjacent second molars, leading to pathological conditions such as external root resorption and dental caries. Accurate assessment of these interactions is important for treatment planning and clinical decision-making. Although cone-beam computed tomography (CBCT) provides detailed three-dimensional imaging, image interpretation remains challenging. Recent advances in artificial intelligence have enabled automated radiographic analysis using deep learning methods. Methods: This retrospective study included 162 CBCT scans obtained from patients aged 18-75 years. A total of 306 third molar-second molar units were evaluated. Based on radiographic findings, interactions were categorized as independent, contact, or resorption. Several deep learning architectures were developed and evaluated, including conventional convolutional neural networks (CNNs), attention-based CNNs, and Vision Transformer (ViT) models. Performance was assessed using standard classification metrics, and an ensemble approach was applied to improve predictive stability. Results: Attention-based and Transformer-based models generally outperformed conventional CNN architectures. These models achieved better discrimination among the defined classes and demonstrated superior overall performance. The ensemble model produced the most reliable results, achieving the highest macro-area under the curve (macro-AUC) values. Distinguishing contact cases from independent cases was the most challenging task, whereas resorption cases were identified more consistently across different models. Conclusions: Transformer-based deep learning models showed promising performance for CBCT-based assessment of third molar-second molar interactions. Ensemble learning further improved classification reliability and robustness. These findings suggest that artificial intelligence-assisted systems may support early detection of third molar-related pathological changes and contribute to more accurate radiological evaluation and clinical decision-making.

Background/Objectives: Novel vision-language models (VLMs) can integrate patient textual data with image data to support medical diagnosis. Recent studies reported conflicting results regarding the performance of multimodal VLMs compared to other models and physician performance. This systematic review aims to assess the diagnostic performance of multimodal VLMs integrating both patient textual and image data across diverse real-world hospital settings. Methods: We performed comprehensive searches of eight resources, including Embase, MEDLINE, and SCOPUS, on 17 December 2025. Eligible studies reporting diagnostic performance of VLMs integrating both image and patient history textual data from real-world adult patients compared to that of other models and physicians were included. The review adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The Prediction model study Risk Of Bias Assessment Tool + AI (PROBAST + AI) was used to assess the quality and risk of bias. The study protocol was registered in the PROSPERO database (CRD420251244054). This review received no external funding. Results: We screened 11,026 records, of which 18 studies met the inclusion criteria. Six studies comparing multimodal and unimodal models demonstrated the consistent superiority of the multimodal models. Four studies evaluating VLM accuracy as standalone agents compared with physician performance reported conflicting evidence. One study assessing VLMs as a clinical copilot demonstrated higher accuracy from the group of physicians using VLM assistance. A meta-analysis could not be performed due to the heterogeneity across study populations and outcomes. The majority of the studies were assessed as having a high risk of bias due to dataset quality. Primary limitations identified across studies include small sample size, a lack of external validation, and the need for prospective clinical deployment studies. No study provided documented considerations regarding model safety or data security. Conclusions: This systematic review suggests that multimodal VLMs consistently outperform unimodal models with access to only image or text. While model performance as standalone agents compared to humans remains inconclusive, a copilot model has demonstrated high diagnostic accuracy. Given substantial methodological concerns across studies, cautious interpretation is required, No firm clinical recommendation can be made regarding the use of standalone VLMs. Further research employing high-quality datasets is needed to ensure the reliability and clinical applicability of future VLMs.

Soccer-video analysis centers on pitch-plane tracking, but camera-view depth cues such as occlusion and goal-area structure are not fully represented on the field plane. Synthetic benchmarks provide dense supervision unavailable for real broadcasts, but whether adaptation yields predictions that are reproducible across matches and operationally feasible remains unclear. We evaluate a Depth Anything V2 model adapted to SoccerNet-Depth with four components: Unaligned MDE accuracy, scale-and-shift aligned diagnostic, match-to-match reliability, and accuracy-cost trade-off. The model achieves an unaligned validation AbsRel of 0.00372. The aligned diagnostic shows that Base DAv2 retained substantial scene-depth structure, whereas SoccerNet adaptation enabled direct compatibility with the normalized target without per-frame ground-truth fitting. Relative to the VKITTI-fine-tuned reference, the adaptation improved all eight metrics in all 21 validation matches, with paired Wilcoxon tests significant after Bonferroni correction. On the challenge split, it reduced AbsRel by 34.1% versus the official baseline. The higher-resolution configuration improved the validation AbsRel by 5.9%, while the default retained a better accuracy-cost balance. At 401.57 ms per frame, the default is suited to post-match analysis, not live or near-real-time use. The study contributes a benchmark-scoped adaptation case study and protocol for foundation MDE on SoccerNet-Depth.

Accurate beef cattle body measurement data are crucial for growth assessment, phenotypic analysis, breeding management, and precision livestock farming. Traditional manual measurements are labor-intensive, time-consuming, and likely to cause stress in animals, making it difficult to meet the demands of large-scale livestock farming. This paper employs a structured systematic literature review method, in accordance with the PRISMA 2020 guidelines, to summarize research progress in vision-based beef cattle body measurement. This paper focuses on reviewing technical approaches such as 2D image-based measurement, 3D measurement using RGB-D and LiDAR, and multi-view fusion. It analyzes key technologies including image segmentation, keypoint detection, point cloud processing, 3D reconstruction, and geometric calculations, and compares the advantages and disadvantages of different methods in terms of measurement accuracy, robustness, cost, and farm applicability. The results indicate that 2D image-based methods are low-cost and flexible to deploy but have limited expressiveness for 3D body measurement parameters; RGB-D and LiDAR methods can provide spatial information but are affected by point cloud noise, occlusion, equipment costs, and data processing complexity; multi-view fusion can improve the completeness of body surface information but places high demands on calibration, registration, and system integration. Current research still faces challenges such as a lack of public datasets, inconsistent annotation standards, uncertainty regarding ground truth, insufficient cross-ranch generalization validation, and limited practical applications. Future research should focus on developing standardized datasets, conducting cross-scenario validation, advancing multimodal perception, creating lightweight models, and applying edge computing to drive the evolution of visual body measurement toward continuous monitoring and intelligent decision-making.

Background/Objectives: Reliable detection of epileptic seizures using electroencephalography (EEG) is crucial for clinical diagnosis and for alleviating clinicians' workload. However, existing studies still make insufficient use of phase information, and the synergy between local time-frequency pattern extraction and global dependency modeling remains limited. Methods: We propose a seizure detection framework based on the continuous wavelet transform (CWT), a three-dimensional convolutional neural network (3D-CNN), and a vision transformer (ViT). First, multichannel EEG segments are preprocessed, after which CWT is used to generate power spectrograms and phase spectrograms. These representations are then fused along the depth dimension into a unified power-phase volume and fed into a hybrid network composed of a 3D-CNN feature extractor and a single-layer ViT encoder to jointly learn local time-frequency-channel coupling patterns and higher-level global dependencies. Finally, seizure detection is completed by combining moving-average filtering, thresholding, and collar correction. Results: On the public CHB-MIT dataset and the clinical SH-SDU dataset, the proposed method achieved average segment-level sensitivities of 98.68% and 92.05%, specificities of 98.33% and 97.53%, accuracies of 98.49% and 96.37%, and AUC values of 97.26% and 92.89%, respectively. In event-level evaluation, the average sensitivities were 99.13% and 96.08%, with false detection rates of 0.88/h and 0.69/h, respectively. Further multi-stage ablation experiments together with t-SNE and Grad-CAM visualizations provided qualitative and experimental support for the design rationale of the joint power-phase input and the hybrid 3D-CNN-ViT architecture. Conclusions: The proposed framework effectively exploits the complementary discriminative value of power and phase information in epileptic EEG and demonstrates strong detection performance under patient-specific evaluation on both public and clinically collected datasets.

Airport turnaround is an important operational process that directly affects flight punctuality, airport capacity, and ground-handling efficiency. However, many turnaround activities are still monitored manually or through fragmented operational records, which can limit real-time visibility and delay identification. This study proposes a computer vision-based airport turnaround monitoring pipeline that integrates YOLOv11 object detection, Norfair multi-object tracking, and frame differencing-based motion analysis to extract key operational events from airport video footage. Publicly available turnaround footage from Shinshu Matsumoto Airport, Japan, was collected under different environmental conditions, including daytime, nighttime, rainy, after-rain, and transition lighting conditions. From selected videos, 1446 images were labeled into 11 airport turnaround object classes, including tow tug, aerobridge, airplane, baggage container, belt loader, belt loader roof, fuel line, fuel tanker, fuel tube, tractor, and window. The dataset was divided into training, validation, and testing sets using a 70:20:10 ratio. The trained YOLOv11 model achieved strong detection performance, with overall test an precision of 0.9609, recall of 0.9445, and mAP50 of 0.9617. To support activity-level interpretation beyond object detection, the proposed pipeline applies frame differencing within specific regions of interest, including the aerobridge window region for passenger deboarding and boarding detection, and the belt loader roof region for baggage unloading and loading detection. The extracted object detections, motion spikes, and temporal logs are then converted into a Gantt chart that summarizes major turnaround activities, including airplane parking, deboarding, baggage unloading, refueling, baggage loading, boarding, and pushback. The results demonstrate that the proposed modified YOLO-based pipeline can transform ordinary airport video footage into structured operational timelines, supporting more transparent, data-driven, and automated monitoring of airport turnaround processes.

Automated mapping of vertical traffic signs and horizontal road markings is essential for road safety and Intelligent Transportation Systems (ITS). Traditional methods are labor-intensive, while existing automated solutions often lack a unified approach or are proprietary, limiting research accessibility and reproducibility. This paper presents a comprehensive framework for identifying these assets using a low-cost, vehicle-mounted action camera. A distance-aware frame extraction strategy is introduced to minimize data redundancy and ensure high spatial diversity. Specific strategies address the class imbalance inherent in real-world driving, ensuring robust detection for infrequent sign categories. Deep learning models handle the distinct geometries of vertical and horizontal assets, employing segmentation-based annotation for irregular road markings. Experimental results show high performance, with leading YOLO-based architectures achieving an F1-score of 0.92 for vertical signage and 0.96 for horizontal markings. By transforming raw visual data into structured georeferenced information, this framework facilitates the generation of High-Definition (HD) maps and digital inventories, supporting road authorities in proactive maintenance planning and regional road safety assessments.

The emergence of multimodal large language models (MLLMs) is opening a new avenue for explainable and interactive intelligent diagnosis in agriculture. However, generic MLLMs still face two major obstacles in plant disease recognition-insufficient fine-grained visual perception and misalignment between visual and linguistic features-which jointly limit diagnostic accuracy. To address these issues, we propose a Qwen2.5-VL-based full-chain fine-tuning framework termed dual-side synergistic low-rank adaptation. Unlike the mainstream paradigm that freezes the vision encoder, our method injects trainable LoRA adapters into both the vision encoder and the large language model, while establishing end-to-end gradient backpropagation across the entire multimodal pipeline. By using the supervision signal from autoregressive text generation (text-supervised visual learning), the framework directly drives deep optimization of visual representations, thereby enabling coordinated alignment between pixel-level perception and semantic-level understanding. We trained Qwen over CDDM and conducted in-domain (CDDM) and cross-domain (PlantVillage) experiments. The results show that the proposed 7B-parameter model achieves 98.8 and 96.0% diagnostic accuracy under in-domain and cross-domain scenarios, respectively. The recognition accuracy of Qwen in the case of cross-domain only decreases slightly, which demonstrates that the MLLM trained by our method exhibits excellent cross-domain recognition capability. This indicates that our method can significantly improve the robustness and generalization ability of MLLM in complex agricultural scenarios.

Background/Objectives: Intermediate vision has become increasingly important after cataract surgery because many daily activities require functional visual performance beyond distance vision alone. Enhanced monofocal intraocular lenses may improve intermediate visual function while maintaining good distance visual acuity. The aim of this study was to compare binocular distance and intermediate visual outcomes after bilateral implantation of Clareon and TECNIS Eyhance intraocular lenses. Methods: This was a single-center, non-randomized comparative observational study with assessment of postoperative visual outcomes. Eighty-six patients who had previously undergone uncomplicated bilateral age-related cataract surgery with implantation of the same IOL model in both eyes were included. Forty-two patients received Clareon IOLs and forty-four patients received TECNIS Eyhance IOLs. Postoperative assessment was performed at least 12 weeks after surgery of the second eye. Binocular corrected distance visual acuity (CDVA) and distance-corrected intermediate visual acuity (DCIVA) were measured using ETDRS charts at 4 m and 66 cm. Results: Baseline biometric and clinical parameters were comparable between groups. Mean postoperative DCIVA was 0.23 ± 0.09 logMAR in the Clareon group and 0.21 ± 0.08 logMAR in the Eyhance group. The mean between-group difference, calculated as Clareon minus Eyhance, was 0.02 logMAR, with a 95% confidence interval from -0.02 to 0.06 logMAR. Mean binocular CDVA was 0.02 ± 0.02 logMAR in the Clareon group and 0.02 ± 0.03 logMAR in the Eyhance group, with a 95% confidence interval for the between-group difference from -0.01 to 0.01 logMAR. Mean postoperative manifest refraction spherical equivalent was -0.27 ± 0.42 D in the Clareon group and -0.29 ± 0.37 D in the Eyhance group. Conclusions: Both Clareon and TECNIS Eyhance IOLs provided good binocular distance and intermediate visual acuity after bilateral implantation. Intermediate visual performance after Clareon implantation was comparable to that achieved with TECNIS Eyhance, while distance visual acuity remained similarly high in both groups. These findings suggest that both IOL models may represent useful options for patients undergoing cataract surgery who expect good distance vision and functional intermediate visual performance.

Ammonia nitrogen is one of the most common environmental stressors in aquaculture water environments, and its accumulation can induce physiological disturbance, altered ventilation regulation, and abnormal behavioral responses in fish. To achieve non-invasive quantitative characterization of rainbow trout responses to ammonia nitrogen stress, this study developed a computer-vision-based framework for the integrated analysis of locomotor behavior and ventilation activity. Rainbow trout were exposed to four ammonia nitrogen concentrations: 0, 15, 30, and 60 mg/L. A total of 16 rainbow trout were used in this study, with an average body length of 14.0 ± 1.0 cm and an average body weight of 38.65 ± 2.42 g. The fish were assigned to four experimental aquaria, with four fish maintained in one aquarium for each TAN treatment. Stereo videos for locomotor behavior analysis and monocular mouth-region videos for ventilation analysis were simultaneously collected, and the final 5 min of each recording was analyzed. YOLOv11n, multi-object tracking, and stereo vision were used to extract three-dimensional position sequences of rainbow trout and calculate the amount of exercise, average swimming speed, and spatial distribution. Meanwhile, optical-flow analysis was applied to quantify mouth opening-closing motion and estimate ventilation frequency. The results showed that with increasing ammonia nitrogen concentration, rainbow trout locomotor behavior tended to be suppressed, with average swimming speed showing the clearest decrease, whereas ventilation frequency continuously increased. Average swimming speed decreased from 3.83 cm/s in the 0 mg/L group to 1.03 cm/s in the 60 mg/L group, while ventilation frequency increased from 84.91 breaths/min to 133.43 breaths/min. Compared with locomotor indicators, ventilation frequency showed a more stable response to changes in ammonia nitrogen concentration. This study achieved the synchronous quantification of rainbow trout locomotor behavior and ventilation activity, revealing a differentiated response pattern characterized by enhanced ventilation and suppressed locomotor behavior under acute ammonia nitrogen stress. These findings provide a methodological reference for fish stress assessment and risk warning in aquaculture environments.

To overcome the limitations of single-dimensional data and low efficiency in traditional cycloidal gear inspection, a comprehensive machine vision-based method was proposed. A high-precision vision platform was established, and a Sigmoid surface-based edge detection algorithm was employed for sub-pixel edge localization. Logarithmic transformation combined with light intensity compensation was applied to correct saturation-induced errors. The pixel equivalent and compensation coefficient were systematically calibrated using a dot-matrix plate and gauge blocks. A sub-pixel tooth profile model in the physical coordinate system was reconstructed through pixel equivalent calibration, dynamic light intensity compensation, and multi-coordinate transformation. Comparative tests against a coordinate measuring machine (CMM) verified that the point-to-point deviation between the two measurement systems was within 10 μm (maximum 11.62 μm). The inherent tooth profile deviation of the tested cycloidal gears, which reflects the machining quality of workpieces, ranged from 24 μm to 37 μm. Multiple repeated tests prove that the system achieves a repeat positioning accuracy of 0.8 μm. Based on the measurement characteristics, a hybrid analytical method integrating Cartesian and polar coordinate systems was developed, enabling the simultaneous evaluation of critical geometric tolerances, such as the diameters of the center hole and crankshaft hole. The full inspection cycle for cycloidal gears was reduced to 13 s, which demonstrates a substantial efficiency improvement over traditional methods.

The integration of collaborative robots into industrial environments requires rigorous safety validation under the Power and Force Limiting (PFL) regime. This review article systematically maps the technological and normative development of certified Pressure and Force Measurement Devices (PFMDs) and experimental biofidelic instruments for Physical Human-Robot Interaction (pHRI) between the years 2011 and 2026. A quantitative screening of 68 studies revealed a publication peak in impact metrology in 2021. This peak occurred with a five-year latency after the release of the ISO/TS 15066 technical specification. Although global interest in collaborative robotics steadily grows, the publication trend indicates a gradual shift in scientific focus from reactive testing toward proactive prevention. A methodological deconstruction of four Research Questions (RQs) identifies persistent limitations in safety evaluation. The findings demonstrate that the internal structure of conventional sensors induces nonlinear shock filtering and parasitic oscillations (RQ1). Furthermore, the rigid fixation of test stands generates unrealistic pressure spikes. This physical limitation forces a transition to flexible and pendulum-based configurations (RQ2). Commercial flat films physically fail due to sensor saturation and introduced stiffness. Such failures accelerate the development of conformable electronic skins (e-skins) and multimodal test manikins (RQ3). To ensure interlaboratory reproducibility within the current ISO 10218-2:2025 standard, the text defines imperative metrological parameters. These parameters strictly include frequency response, calibration protocols, and volumetric mapping of inertial masses (RQ4). Furthermore, the analysed publications were systematically stratified into distinct technological categories, strictly reflecting their primary engineering domains, ranging from empirical metrological evaluation and sensor hardware design to advanced numerical modeling. Finally, the vision for future research anticipates a definitive shift toward proactive anti-collision technologies, encompassing Artificial Intelligence (AI), machine vision, and Augmented Reality/Virtual Reality/Mixed reality (AR/VR/MR). Future methodologies must also consider demographic anisotropies and the cognitive fatigue of the human operator.