Accurate crack assessment in electroluminescence (EL) images is important for photovoltaic (PV) reliability analysis, yet existing segmentation methods often fail to capture the thin, elongated, and structurally constrained nature of crack defects. This paper proposes a Solar Topology Crack Network (STC-Net) that incorporates edge priors, spectral priors, and a boundary-topology refinement module to improve crack continuity and boundary preservation.
Automated solar filament detection using deep learning faces several challenges. Semantic segmentation of solar filaments is a complicated multiscale feature extraction task with long-tail distribution.
The paper proposes a hybrid approach for detecting surface defects on solar panels by combining handcrafted features—Local Binary Pattern, Histogram of Gradients, and Gabor Filters—with deep features extracted from DenseNet‑169. The concatenated feature set is classified using SVM, XGBoost, and LGBM, with the DenseNet‑169 + Gabor (SVM) configuration achieving the highest accuracy of 99.17% on an augmented dataset. This method aims to provide an automated, accurate, and flexible defect‑detection system for large‑scale, remote solar power plants.
The paper presents a reproducible workflow built in the Galaxy environment for automated crack identification and damage assessment from scanning electron microscopy images of fusion materials. It processes SEM images and metadata to generate crack masks, skeletonized networks, quality‑control visualizations, and scalar damage descriptors without needing image‑specific tuning across different tungsten grades, microstructures, magnifications, and damage states. The workflow was demonstrated on a dataset of 418 images from 114 experiments covering five tungsten grades and three microstructural states, and it includes a crack‑density descriptor that serves as a standardized input for downstream machine‑learning prediction and physics‑based crack simulation modules.
arXiv:2609.16773v1 Announce Type: new Abstract: Image segmentation remains challenging due to occlusions, poor lighting, and irregular structures. Although transformer-based methods achieve high accu...
Image segmentation remains challenging due to occlusions, poor lighting, and irregular structures. Although transformer-based methods achieve high accuracy, they rely heavily on long-range spatial fea...
The paper introduces a mesh‑free multiresolution deep energy method for phase‑field modeling of brittle fracture. A single neural network represents displacement and phase fields, trained by minimizing incremental energy with multiresolution B‑spline feature encoding and stratified Monte Carlo integration. Across six benchmark problems, the method reproduces load‑displacement curves and crack patterns with high accuracy, outperforming a deep Ritz baseline on a random multi‑crack dataset.
The paper introduces an end‑to‑end automated pipeline that generates controllable crack data for deep‑learning inspection. It uses procedurally sampled Bézier‑curve skeletons converted into realistic crack masks via a GAN, and a dual‑ControlNet diffusion model that separates appearance from geometry while enforcing boundary consistency. The method supports both background‑free synthesis and context‑aware inpainting, and shows improved performance over existing augmentation baselines on CRACK500 and CrackTree200 datasets.
The paper introduces YOLOEZ, a no-code, GUI-based tool that streamlines the entire YOLO model workflow—data labeling, training, and inference—for automated structural defect detection. It demonstrates that YOLOEZ outperforms traditional image‑processing methods across most detection metrics while simplifying deployment for users without programming expertise. The tool aims to lower technical barriers in structural health monitoring, enabling broader adoption of AI-driven inspection for predictive maintenance and intelligent structural systems.
Phase-field modeling of brittle fracture removes the need to track cracks explicitly by recasting their evolution as the minimization of an energy functional. In return it requires a discretization de...
arXiv:2607. 08794v1 Announce Type: cross Abstract: Sand boils on earthen levees are safety-critical defects, but pixel-level detection is limited by scarce annotations.
The paper presents a data‑efficient method for quantifying cracks in lithium‑ion cathodes using a frozen self‑supervised vision‑transformer encoder, a lightweight decoder, and iterative model‑assisted annotation. Applied to three 120‑megapixel NMC cathode cross‑sections, the framework distinguishes intragranular from early and late intergranular cracks, providing per‑particle distributions of crack width, tortuosity, and area fraction. Late intergranular crack coverage reaches 4.6% in cycled samples versus 0.5% in initial and calendar‑aged samples, indicating degradation primarily from electrochemical cycling.
The paper introduces RABR-Net, a two‑stage framework that refines biomedical image segmentation boundaries by combining multiple uncertainty measures into a boundary‑aware representation. A gated residual refiner uses this representation to selectively correct uncertain boundary pixels while preserving confident regions, leading to modest but statistically significant improvements in Dice, Boundary Dice, and HD95 metrics on a held‑out test set. Qualitative results show the refiner focuses on uncertain cytoplasm and nucleus boundaries, though calibration does not automatically improve.