Search Results (2,001)

The objective of this study was to evaluate the accuracy of detecting crop damage caused by wild boar in rapeseed fields using UAV (unmanned aerial vehicle)-derived RGB (red, green and blue) imagery and deep learning segmentation models. The experiments were conducted on rapeseed crops at full maturity shortly before harvest in central-western Poland in 2021. Four convolutional neural network architectures—U-Net (U-shaped network), U-Net++, DeepLabV3+ (deep learning + labelling), and PSPNet (Pyramid Scene Parsing Network)—were benchmarked using two input configurations: RGB imagery alone and RGB combined with the topographic position index (TPI) derived from a digital surface model (DSM). Model performance was assessed using overall accuracy, F1-score (harmonic mean of precision and recall), and Intersection over Union (IoU), with class-specific metrics reported to provide a realistic evaluation of damaged-area detection. For RGB-only data, overall accuracy ranged from 0.957 to 0.972, while damaged-class F1 and IoU reached 0.752 and 0.603, respectively, for the best-performing model (U-Net). When RGB data were supplemented with TPI, overall accuracy and damaged-class metrics changed only slightly, indicating limited benefit from the topographic feature under these field conditions. Non-damaged crop areas were consistently well-classified (F1 > 0.977, IoU > 0.955). These results confirm that UAV-based RGB imagery enables reliable late-season assessment of wildlife-induced crop damage, and that reporting class-specific metrics in spatially independent test sets is essential for realistic performance evaluation. Full article

Remote sensing-based Earth observation provides essential spatial data for analyzing and monitoring both natural and urban environments. Precise characterization of objects in these scenes is vital for environmental management, land-use planning, and monitoring global change. Semantic segmentation of remote sensing imagery (RSI) is a fundamental yet complex task due to significant variability in object shape, scale, and distribution, as well as the complexity of multiscale landscapes captured by advanced sensors. Convolutional neural networks, especially the U-Net architecture, have achieved notable success in segmentation tasks. However, their application in remote sensing is often impeded by persistent issues such as loss of spatial detail, substantial intra- and inter-class variability, and high sensitivity to hyperparameter settings. Manual tuning of hyperparameters is typically inefficient and error-prone, which highlights the importance of heuristic methods for automated optimization. Genetic Algorithms (GAs), Differential Evolution (DE), and Particle Swarm Optimization (PSO) are metaheuristics that provide systematic approaches for exploring large hyperparameter spaces. This study investigates an evolutionary framework for the automated optimization of four critical U-Net hyperparameters—learning rate, number of training epochs, optimizer, and loss function—using micro-evolutionary algorithms. Specifically, micro Genetic Algorithms (micro-GAs), micro Differential Evolution (micro-DE), and micro Particle Swarm Optimization (micro-PSO) are employed to efficiently explore the hyperparameter search space under reduced population settings. The experimental results demonstrate that the proposed micro-evolutionary optimization framework consistently enhances segmentation performance, achieving improvements in Mean Intersection over Union (MIoU) ranging from 3% to 35%, along with systematic gains in overall accuracy across different datasets and configurations. Full article

Grain breakage occurs in sand specimens subjected to high stress levels; however, the magnitude and characteristics of the resulting deformation remain insufficiently quantified. This study investigates particle-scale fracture behavior in a standardized quartz sand subjected to controlled mechanical loading. Rapid, unconsolidated–undrained (UU) direct shear box tests were performed under normal stresses of 700, 800, and 900 kPa to induce grain breakage. The mechanical loading procedure was applied as a controlled stress induction mechanism to promote particle fragmentation rather than to determine conventional geotechnical parameters. A uniformly prepared quartz sand containing no additional mineral phases was used to ensure material consistency. Post-test specimens were examined through systematic visual and image-based analysis. The sample obtained from the 900 kPa test, where breakage was most pronounced, was analyzed in detail to characterize quartz fracture behavior under compressive and shear stress conditions using advanced image processing techniques. A deep learning-based mineral segmentation framework was developed using a ResNet50 architecture with transfer learning. A custom dataset consisting of high-resolution mineral images and corresponding pixel-level segmentation masks was constructed. The proposed model achieved 86.21% overall accuracy, a Dice coefficient of 91.35%, and an Intersection-over-Union (IoU) score of 84.07%. Validation results demonstrated strong generalization capability, with validation accuracy, Dice score, and IoU of 87.47%, 90.07%, and 81.96%, respectively. The high-precision segmentation performance enabled a comprehensive fracture analysis of 3333 quartz mineral images obtained from specimens exposed to systematic stress conditions. Full article

This research significantly improves the link reliability and robustness of optical camera communications (OCC) by leveraging deep learning for light source modulation filtering, reflection filtering, and precise light source localisation. By using image sensors as receivers in OCC, data transmission is not only enabled, but other applications are also facilitated, such as detecting objects and humans, making OCC highly attractive in healthcare, intelligent transport systems, and indoor positioning. However, the position of the desired signal in the received image frame must be tracked in dynamic scenarios (i.e., nonstationary applications), in order to maintain the communication link. Moreover, as sixth-generation (6G) wireless networks envision highly autonomous systems that rely on seamless integration of communication and sensing, deep learning is key to enabling robust and adaptive light source localisation and sensing in OCC, which enables vision-based autonomy in dynamic environments. It should be noted that a deep learning-based approach provides more accuracy even when there are multiple noise sources in the environment, reflections, and complex backgrounds, and under mobility conditions, in which traditional light source detection/tracking methods are not effective. Hence this study investigates the use of a deep learning-based approach by analysing the detection accuracy under different configurations and unseen images. The results obtained demonstrate consistently high detection performance with average precision (at an intersection-over-union threshold of 0.70 of 0.84 to 0.97. These results pave the way for autonomous receivers that will be able to select signals intelligently and decode them. Full article

This study proposes a clinically driven hybrid segmentation framework for dense breast tissue analysis in mammographic images, addressing persistent challenges associated with intensity inhomogeneity, low-contrast, and complex tumor morphology. The framework integrates Multiplicative Intrinsic Component Optimization (MICO_2D) for bias field correction, followed by a distance-regularized multiphase Vese–Chan level-set model for coarse global tumor segmentation. To achieve precise boundary delineation, a localized refinement stage is employed using Localized Active Contours (LAC) with Local Image Fitting (LIF) energy, supported by Gaussian regularization to ensure smooth and coherent boundaries in regions with ambiguous tissue transitions. Building upon the refined semantic tumor mask, the framework further incorporates a panoptic-style tumor instance segmentation stage, enabling the decomposition of connected tumor regions into distinct anatomical instances, which were evaluated on both MIAS and INBreast mammography datasets to demonstrate generalizability. This extension facilitates detailed structural analysis of tumor multiplicity and spatial organization, enhancing interpretability beyond conventional pixel wise segmentation. Experiments conducted on Cranio-Caudal (CC) and Medio-Lateral Oblique (MLO) mammographic views demonstrate competitive performance relative to baseline U-Net and advanced deep learning fusion architectures, including multi-scale and multi-view networks, while offering improved interpretability and robustness. Quantitative evaluation using overlap-related metrics shows strong spatial agreement between predicted and reference segmentations, with per-image Dice Similarity Coefficient (DSC) and Intersection over Union (IoU) distributions reported to ensure reproducibility. Descriptive per-image analysis, supported by bootstrap-based confidence intervals and paired comparisons, indicates consistent performance improvements across images. Robustness analysis under realistic perturbations, including noise, contrast degradation, blur, and rotation, demonstrates stable performance across varying imaging conditions. Furthermore, feature space visualizations using t-SNE and UMAP reveal clear separability between cancerous and non-cancerous tissue regions, highlighting the discriminative capability of the proposed framework. Overall, the results demonstrate the effectiveness, robustness, and clinical motivation of this hybrid panoptic framework for comprehensive dense breast tumor analysis in mammography, while emphasizing reproducibility and conservative statistical assessment. Full article

Surface defects in industrial production are complex and diverse. Therefore, deep learning-based defect detection models must consistently adapt to newly emerging defect categories. The trained models generally suffer from catastrophic forgetting as they learn new defect categories. To address this issue, we propose a selective distillation for incremental defect detection (SD-IDD) model based on GFLv1. Specifically, three selective distillation strategies are proposed, including high-confidence classification distillation, dual-stage cascaded regression distillation, and Intersection over Union (IoU)-driven difficulty-aware feature distillation. The high-confidence classification distillation aims to preserve critical discriminative knowledge of old categories within semantic confusion regions of the classification head, reducing interference from low-value regions. Dual-stage cascaded regression distillation focuses on high-quality anchors through geometric prior coarse filtering and statistical fine filtering, utilizing IoU-weighted KL divergence distillation loss to accurately transfer localization knowledge. IoU-driven difficulty-aware feature distillation adaptively allocates distillation resources, prioritizing features of high-difficulty targets. These selective distillation strategies significantly mitigate catastrophic forgetting while enhancing the detection accuracy of new classes, without requiring access to old training samples. Experimental results demonstrate that SD-IDD achieves superior performance, with mAP_old of 58.2% and 99.3%, mAP_new of 69.0% and 97.3%, and mAP_all of 63.6% and 98.3% on the NEU-DET and DeepPCB datasets, respectively, surpassing existing incremental detection methods. Full article

Background/Objectives: Hepatocellular carcinoma (HCC) remains a leading cause of cancer-related mortality worldwide, where accurate liver and tumor segmentation from magnetic resonance imaging (MRI) is essential for diagnosis, treatment planning, and disease monitoring. Despite recent advances, MRI-based segmentation remains challenging due to data heterogeneity and limited annotated datasets. This study aims to systematically compare convolutional, transformer-based, and detection-based deep learning approaches for liver and HCC segmentation using contrast-enhanced MRI. Methods: A comprehensive evaluation was conducted on the ATLAS MRI dataset, including 2D- and 3D-CNN, transformer-based architectures, and single-stage YOLO-based segmentation frameworks. All models were trained using consistent preprocessing, patient-level data splits, and standardized evaluation metrics, including Dice coefficient, Intersection over Union (IoU), precision, recall, and F1-score. Results: Volumetric convolutional models achieved the highest segmentation accuracy, with the 3D nnU-Net yielding superior performance for both liver (Dice: 0.946) and tumor (Dice: 0.892) segmentation. Transformer-based models demonstrated competitive results, particularly in capturing global contextual information and improving boundary delineation, while YOLO-based approaches provided balanced accuracy with substantially reduced computational cost. Conclusions: The findings confirm that volumetric CNNs remain the most accurate solution for MRI-based liver and HCC segmentation, whereas transformer- and YOLO-based frameworks offer complementary advantages for specific clinical and resource-constrained scenarios. Full article

Background: Bone Mineral Density (BMD) plays a crucial role in diagnosing osteoporosis, and early detection is essential to preventing complications such as osteoporotic fractures. However, access to dual-energy X-ray absorptiometry (DXA) screening remains limited in many healthcare settings. Objective: This study presents a fully automated artificial intelligence pipeline for BMD prediction from lumbar spine radiographs to enable opportunistic osteoporosis screening. Methods: The proposed system integrates automatic vertebral segmentation and a machine learning-based regression model for BMD prediction. A YOLO-based instance segmentation model was trained to automatically segment four lumbar vertebrae, achieving a high Intersection over Union (IoU) of 0.9. Radiomic features were extracted from the segmented vertebrae to capture advanced image characteristics and combined with clinical features from 2875 female patients. An eXtreme Gradient Boosting (XGBoost) regressor was trained to provide opportunistic BMD estimation. Results: The model achieved a mean absolute percentage error (MAPE) of 6% for BMD prediction. A classification model built from segmented vertebrae distinguished between osteoporosis, osteopenia, and normal bone with approximately 90% accuracy. Strong agreement between predicted and ground-truth BMD values was confirmed using Pearson correlation coefficient and Bland–Altman analysis. Conclusions: The proposed fully automated system demonstrates strong agreement with DXA measurements and potential for opportunistic osteoporosis screening in settings with limited DXA access. Further validation and refinement are needed to achieve clinical-grade precision for diagnostic applications. Full article

To address the challenges of thin cloud detection and imprecise cloud boundary segmentation in Landsat 8 remote sensing imagery, this paper proposes a systematic approach that comprehensively enhances cloud detection accuracy from data preprocessing to network architecture optimisation. First, through empirical analysis, an optimised band input combination was determined (removing the panchromatic Band 8 and thermal infrared Band 11), effectively suppressing urban background noise. Subsequently, an enhanced ResUNet model was designed, innovatively integrating an Atrous Spatial Pyramid Pooling (ASPP) module with an attention gate (AG) mechanism. The ASPP module enhances detection capabilities for thin clouds and diffuse cloud masses by aggregating multi-scale global contextual information. The attention-gated mechanism finely tunes feature fusion during the decoding phase, suppressing interference from highly reflective surface features to achieve precise cloud boundary segmentation. Experiments conducted on the Landsat 8 dataset featuring typical urban scenes demonstrate that the proposed method significantly outperforms mainstream models across both conventional and boundary-specific metrics, achieving an overall accuracy (OA) of 0.9717, a mean intersection over union (mIoU) of 0.8102, and, notably, a mean bounding box intersection over union (mB-IoU) of 0.4154 and a mean bounding box F1 score of 0.5356, representing improvements of 16.3% and 12.5%, respectively, over existing methods. This research provides an efficient and robust technical framework for cloud detection tasks in complex urban environments, laying the foundation for high-precision processing of remote sensing imagery and subsequent quantitative analysis. Full article

Given the rising global demand for environmentally sustainable energy sources, solar photovoltaic (PV) power generation has emerged as a pivotal component of the energy transition. In PV systems, power conversion efficiency is degraded and operational lifespan reduced due to the presence of defective modules. Consequently, achieving accurate and efficient defect detection during PV module manufacturing is critical to ensuring product quality and reliability. To address this challenge, we propose MambaVSS-YOLOv11n, an electroluminescence (EL) image-based multi-defect detection method for PV modules. Our study utilizes a dataset containing six types of defects—Broken Gate, Cold Solder Joint, Black Spot, Scratch, Microcrack, and Suction Mark—to construct 692 labeled EL images of defective PV modules. The model integrates the Vision State Space (VSS) module from Mamba and optimizes the C3k2 Bottleneck structure to enhance fine-grained feature extraction, while employing Space-to-Depth Convolutional (SPD-Conv) Layer for downsampling to improve computational efficiency. Additionally, to address YOLOv11n’s limited generalization capability for small objects and complex backgrounds, we adopt the Inner Mask Distance Penalized Intersection over the Union (Inner-MDPIoU) loss function, which enhances detection accuracy and mitigates the impact of low-quality samples. Experimental results demonstrate that compared to YOLOv11n, MambaVSS-YOLOv11n reduces the number of parameters by 18.1%, while improving mAP@0.5 to 0.869 and mAP@0.5:0.95 to 0.637. This achieves model lightweighting while enhancing detection performance. These findings indicate that the model is well-suited for real-time defect detection in PV module production lines, providing PV manufacturers with a lightweight yet accurate and reliable solution for PV module defect inspection. Full article

Cropland constitutes a foundational resource for global food security and agricultural sustainability, and its accurate extraction from high-resolution remote sensing imagery is essential for agricultural monitoring and land management. However, existing deep learning-based segmentation methods often struggle to balance global contextual modeling and fine-grained boundary representation, leading to boundary blurring and omission of small cropland parcels. To address these challenges, this study proposes a novel CNN–Transformer dual-branch fusion network, named CAFM-Net, which integrates a convolution and attention fusion module (CAFM) and an edge-assisted supervision head (EH) to jointly enhance global–local feature interaction and boundary delineation capability. Experiments were conducted on a self-built PlanetScope cropland dataset from Suihua City, China, and the GID public dataset to evaluate the effectiveness and generalization ability of the proposed model. On the self-built dataset, CAFM-Net achieved an overall accuracy (OA) of 96.75%, an F1-score of 96.80%, and an Intersection over Union (IoU) of 93.79%, outperforming mainstream models such as UNet, DeepLabV3+, TransUNet, and Swin Transformer by a clear margin. On the GID public dataset, CAFM-Net obtained an OA of 94.58%, an F1-score of 94.19%, and an IoU of 89.02%, demonstrating strong robustness across different data sources. Ablation experiments further confirm that the CAFM contributes most significantly to performance improvement, while the EH module effectively enhances boundary accuracy. Overall, the proposed CAFM-Net provides a quantitatively validated and robust solution for fine-grained cropland segmentation from high-resolution remote sensing imagery, with clear advantages in boundary precision and small-parcel detection. Full article

Semantic segmentation is a critical task in computer vision, enabling dense classification of image regions. This work investigates the effectiveness of the DeepLabv3+ architecture for binary semantic segmentation using annotated image data. A pretrained ResNet-101 backbone is employed to extract deep features, while Atrous Spatial Pyramid Pooling (ASPP) and a decoder module refine the segmentation outputs. The dataset provides per-image annotations indicating class presence, which are leveraged to approximate segmentation masks for training purposes. Various data augmentation techniques and training strategies were applied to support effective learning and reduce overfitting. Experimental results on the MHIST dataset show that the proposed pipeline achieves strong performance despite the lack of pixel-level annotations, with a mean Intersection-over-Union (mIoU) of 0.76 and a mean Dice coefficient of 0.84. These confirm the potential of weakly supervised segmentation using class-aware CAMs and deep pretrained encoders for structured pixel-level prediction tasks in medical imaging. Full article

Cloud detection in remote sensing imagery is a research hotspot in the field of image processing, and accurately detecting and segmenting cloud regions is crucial for improving the utilization efficiency of remote sensing data. However, standard convolutional neural networks face limitations in modeling the complex spatial structures of clouds. To address these challenges, this paper proposes a cloud detection method based on DDS-DeeplabV3+. First, a lightweight design of the Xception network is adopted to control model complexity, and part of its standard convolutional layers are replaced with Deformable Convolutional Networks (DCN), which enhances the capability of the model to capture geometric features of irregular cloud formations. Second, a Dual-Branch Collaborative Mechanism (DCM) that integrates global context modeling with local detail perception is designed to reconstruct the Atrous Spatial Pyramid Pooling (ASPP) module, thereby improving performance in handling complex scenes and fine boundary delineation. Finally, the SimAM (Simple, Parameter-Free Attention Module) is incorporated into the decoder module, enhancing thin cloud detection capability. Experimental results on the Landsat-8 and GF-1 datasets show that the proposed model achieves Mean Intersection over Union (MIoU) values of 92.61% and 94.04%, respectively, outperforming other comparative methods and demonstrating its superior performance in cloud detection tasks. Full article

Plastic mulching is widely used in arid and semi-arid cotton systems to improve soil hydrothermal conditions and water–nutrient use efficiency. However, residual mulch and its potential contribution to microplastic inputs pose growing environmental and soil-quality risks, highlighting the need for high-resolution and automated approaches to support plastic waste management, targeted retrieval, and precision field operations. Taking a mulched cotton field in Alar, Xinjiang, as the study area, this study proposes a novel plastic mulch extraction method that integrates Unmanned Aerial Vehicle (UAV)-based hyperspectral imagery with deep learning semantic segmentation. The Jeffries–Matusita (JM) distance was employed to select highly separable optimal bands and their combinations for discriminating plastic mulch, bare soil, and cotton canopy, which were then used to drive UNet, DeepLabV3+, and PSPNet models for plastic mulch mapping. The results indicate that the PSPNet model driven by the 402 nm single-band reflectance, Normalized Difference Index (NDI) (861 nm, 410 nm), and NDI (757 nm, 676 nm) achieved the best performance for plastic mulch identification (Intersection over Union (IoU) = 80.28%), significantly outperforming the RGB-based model (IoU = 76.51%). This study enables accurate, spatially explicit assessments of residual mulch, providing actionable evidence for plastic waste monitoring and management, while supporting sustainable agriculture and precision farmland management. Full article

During long-term service, concrete structures are exposed to various adverse factors, which often lead to the formation of numerous surface cracks. These cracks pose serious threats to structural safety and durability. Therefore, accurately identifying crack characteristics is essential for evaluating the service performance of concrete structures. A two-stage concrete crack segmentation method is presented in this study. The crack is initially located by the improved YOLOv11 that integrates three novel modules, namely Multi-scale Edge Information Enhancement, Efficient-Detection, and P2-Level Feature Integration, to form the MEP-YOLOv11 model. Then, the detected region is taken as input prompts for Segment Anything Model (SAM) to achieve precise crack segmentation. This approach eliminates the need for manual prompting in SAM, enabling automatic crack feature identification. The average Accuracy, precision, and Intersection over Union (IoU) for crack segmentation are 95.98%, 92.60%, and 0.77, respectively. To further enhance the robustness of the two-stage segmentation method under non-uniform illumination conditions, a mask re-input strategy is introduced. The crack mask generated by SAM using bounding-box prompts is fed back into SAM to guide a second round of segmentation. Experimental results demonstrate that the improved method maintains high segmentation performance, with an average Accuracy of 92.38%, precision of 85.70%, and IoU of 0.64. Overall, the proposed method meets engineering requirements for high-precision and efficient crack detection and segmentation, showing strong potential for practical inspection tasks. Full article