Search Results (13,703)

Background/Objectives: Schatzker type is assigned early, usually from the anteroposterior (AP) radiograph. A single benchmark accuracy cannot say whether a model read the fracture, the anatomy around it, the annotation, or how the archive was assembled. We ran four inexpensive controls to separate those contributions. Methods: We benchmarked a ResNet-50 on PlaTiF, a 2026 public release built for artificial-intelligence research that pairs 421 AP knee radiographs from 186 patients with expert Schatzker labels and per-image tibial segmentations. Evaluation used stratified group five-fold cross-validation grouped by patient, five seeds and balanced accuracy. Inputs were cropped to the expert tibial segmentation shipped with the dataset, an oracle localisation unavailable at deployment. Four controls ran on identical folds: a regression given no pixel content; ablation of the tibial pixels with its complement; a regression on the expert mask alone; and an augmentation audit for label-erasing invariances. Results: Among the 128 fracture patients the network reached 0.345 ± 0.030 six-class balanced accuracy, +0.168 over a non-anatomical baseline fitted on the same folds and the same labels (95% CI +0.106 to +0.230, p = 0.002). Recall was graded: 0.72 for Schatzker VI, 0.11 for V and 0.04 for IV, the last two below chance (0.167). Erasing the tibial pixels left 0.257 ± 0.025, read on its own as the target bone being unused; its complement, the tibia with everything else removed, reached 0.367 ± 0.012, and the whole radiograph, which carries both, only 0.297 ± 0.032 (+0.071 for the tibia alone, 95% CI +0.031 to +0.111, p = 0.008). A regression on the expert mask alone reached 0.213 ± 0.034 and was not distinguishable from the erased model. On fracture versus no classifiable fracture the network reached 0.833 ± 0.028 against 0.814 ± 0.016 for a model given no pixels (p = 0.264), and a coronal computed tomography section accompanied 126 of 128 fracture patients but 24 of 58 others (p = 2.9 × 10⁻¹⁹). Conclusions: Each headline number admitted an explanation other than the fracture in the target bone. An ablation reported without its complement misstated where the signal lay, and an augmentation audit overturned our own explanation for the failure of type IV. Controls of this kind cost minutes, and this study illustrates why they can be informative when a benchmark is built on a retrospective clinical archive. Full article

This study investigated the long-term relationship between land-use change and land-cover-based ecological quality patterns in Ordos City, a typical coal resource-based city in northern China. To explicitly capture land-use transitions driven by coal exploitation, mining areas were classified as an independent land-cover type. An improved U-Net semantic segmentation model integrating multispectral information and land surface temperature was subsequently employed to generate multi-temporal land-cover maps. In the internal semantic segmentation validation, the proposed model achieved a mean Intersection over Union (mIoU) of 69.51%, a mean accuracy (mAcc) of 81.65%, and a pixel-level overall accuracy (aAcc) of 83.16%. An independent point-based accuracy assessment of the final land-cover map yielded an overall accuracy (OA) of 66.00% and a Kappa coefficient of 0.6033, indicating acceptable classification reliability in complex mining areas. Based on the classification results, three indicators, namely the land-use transition matrix, ecological environmental quality index (EQI), and ecological contribution index, were adopted to analyze spatiotemporal land-use dynamics and associated land-cover-based ecological quality patterns in Ordos City over the past 25 years. The results indicate that: (1) Marked land-use changes occurred in the land-use pattern of the study area during 2000–2025, with grassland and unused land consistently remaining the dominant land-use types. The area classified as Mining increased from 105.83 km² to 1184.17 km² in 2020, exhibiting distinct characteristics of phased expansion and subsequent adjustment. Built-up land continued to expand, whereas the water area decreased by 47.55%. (2) The land-cover-based EQI exhibited a pattern of decline, recovery, and stabilization, decreasing from 0.529 in 2000 to 0.489 in 2015 before recovering to 0.522 in 2025. This pattern was associated with changes in land-cover composition over the study period, including mining expansion and restoration-related transitions. The findings provide a scientific basis for ecological restoration planning and high-quality transformation in Ordos and other coal resource-based cities with similar arid and semi-arid environmental conditions. Full article

Perceived safety determines whether urban greenways are used in everyday life, yet it is rarely measurable at the boundary scale where design decisions are made. Existing street-view models split into black-box networks whose predictions cannot be traced to design elements and pixel-ratio regressions whose interpretability rests on weak, unstructured representations, while greenspace studies lean on GIS proximity variables that confound design with context. We present the CPTED-Guided Perception Network (CGPN), which fuses a visual branch with a masked, learnable projection of segmentation ratios onto five CPTED dimensions. Because the mask confines learning to a theory-defined support, the prior regularizes the representation while every coordinate of the model remains tied to a named CPTED dimension, whose directional effect on the prediction we verify by perturbation. On 110,633 street-view images, CGPN is statistically equivalent to the strongest black-box baseline in pairwise ranking accuracy (0.649 vs. 0.652; equivalence test within a 1.5-point margin, $p=0.006$, attains the best $R^2$ (0.192), and improves on its unconstrained variant in goodness of fit across three seeds ($\mathrm{\Delta }{R}^2=+0.031$, $p=0.042$). Applied to 218 greenway-adjacent residential boundaries in Boston and New York, it uncovers a threshold-like negative association for barrier-dominated access control and an inverted-U distance profile whose weakest segment lies within 100 m of the greenway edge ($p=0.007$). Full article

Estimating metric depth for safety-relevant targets from a top-view crane camera is challenging because monocular predictions are scale-ambiguous and target regions of interest (RoIs) mix returns from payloads, personnel, hoisting wires, and the ground. We present a target-focused pipeline in which a detector defines RoIs, a dense metric-depth map is predicted, and one representative depth is extracted per object. A DINOv2–DPT relative-depth network is fixed, while a multiscale adapter, DINO detection branch, and pixel-wise spatial affine calibration head are trained. Ground-truth bounding boxes restrict metric supervision to target-relevant regions but are not inputs to the calibration head. At evaluation, all depth methods use identical predicted RoIs, and a class-aware kernel-density estimation rule extracts representative depths. On 272 hardware-synchronized RGB–LiDAR frames from one operational 150-ton crane, the proposed method achieved a target-level RMSE of 0.564 m and a pixel-level inlier ratio of 0.966 for $\delta <1.25$. One-to-one ground-truth matching yielded a 0.566 m target RMSE with detection precision/recall of 0.960/0.970. AdaBins obtained lower pixel-level MAE and RMSE, whereas the proposed method obtained lower target-level errors. These results demonstrate the feasibility of target-focused metric ranging under the evaluated operational crane conditions. Full article

Hyperspectral image classification (HSI) requires a model to distinguish subtle spectral differences while preserving the spatial structure of land-cover regions. CNN-based methods are effective for local spectral–spatial extraction, but their limited receptive fields can weaken broader context modelling. Transformer-based methods improve long-range dependency modelling, yet fixed patch partitioning may reduce their sensitivity to fine local structures. To address these limitations, this study proposes the Dual-Stream Spatial–Spectral Network with Nested Attention (DSSN), which separates local spectral–spatial feature extraction from multi-scale spatial-context modelling before adaptive fusion. The DSSN combines a cascaded 3D-CNN spectral stream, a nested Transformer spatial stream with pixel-level and patch-level interactions, and a channel-attention-based adaptive fusion module. Experiments on Indian Pines, Pavia University and Salinas show DSSN achieves overall accuracies of 98.11%%, 99.88% and 99.82%, respectively, outperforming other baselines. The ablation experiments confirm that each major component contributes to the final performance. Although the model requires more parameters and longer inference time than several compared baselines, its inference time remains at the millisecond level. These results suggest that decoupled spatial–spectral representation and adaptive multi-scale fusion can improve hyperspectral image classification under the evaluated benchmark settings. Full article

Secure transmission and storage of color images remain challenging tasks due to strong inter-pixel correlations and high data volume. This work proposes a one-dimensional sine-tent-logistic-exponential map (STLEM) equipped with numerical boundary correction rules to mitigate finite-precision numerical degradation so as to enhance the unpredictability of chaos-driven cryptosystems. We benchmark STLEM against classic logistic, tent, and sine maps via Lyapunov exponents, autocorrelation, approximate entropy, permutation entropy, Lempel-Ziv complexity, and Kolmogorov–Sinai entropy. Bifurcation diagrams, the 0–1 test, and NIST statistical tests are further adopted to characterize its chaotic dynamics and randomness. Comparative results verify that STLEM achieves improved dynamical complexity and randomness performance. Built upon the proposed STLEM, this paper constructs a color-image encryption scheme that employs a 256-bit master key and two groups of chaotic parameters to produce key-related chaotic sequences. The cryptosystem integrates dynamic edge expansion, chaotic permutation, position-dependent adaptive DNA encoding, DNA-domain chained diffusion, and two successive row-column permutation phases. HMAC-SHA-256 is utilized to generate plaintext-aware initial conditions and perform ciphertext authentication prior to decryption. Experimental validations demonstrate complete plaintext recovery under valid secret inputs, while authentication rejects invalid keys and tampered ciphertexts. Ciphered images exhibit high information entropy, negligible adjacent-pixel correlations, and satisfactory number of pixel change rate (NPCR) and unified average changing intensity (UACI) metrics. Benefiting from a sufficiently large key space and $O\left(MN\log \right(MN\left)\right)$ computational complexity, the proposed scheme is resilient against brute-force attacks and well suited for secure color-image communication scenarios, rather than acting as a general-purpose replacement for standard block ciphers. Full article

Mirror segmentation aims to identify mirror pixels from a single RGB image, yet remains challenging because mirrors provide weak intrinsic texture cues and their appearance is dominated by scene-dependent reflections under varying illumination and viewpoints. While recent models improve performance by leveraging contextual contrast, symmetry priors, frequency/spectral cues, or additional modalities (e.g., depth), many cross-cue or symmetry-aware designs still rely on direct spatial-domain fusion, such as concatenation, addition, or attention. Such fusion can amplify reflection-induced high-frequency variations and lead to leakage, shape distortion, and unstable boundaries. In this paper, we propose a symmetry-aware mirror segmentation framework that stabilizes cross-branch interaction via a frequency-domain cross-conditioned fusion mechanism. We build a dual-path Siamese encoder using the original image and its horizontally flipped counterpart, and introduce Heat Conduction Operator-based Cross Fusion (HCOCF), which performs heat-conduction-inspired spectral attenuation in the DCT domain. Unlike conventional fusion, HCOCF generates a nonnegative cross-conditioned attenuation coefficient map from the opposite branch and applies it to the DCT coefficient grid of the target branch. This produces a DCT-domain attenuation mask that controls the spectral refinement strength of each target feature stream, enabling global context propagation while suppressing unstable reflection-induced high-frequency responses without aggressive direct feature mixing. For multi-scale decoding, we adapt the cross-scale decoder of the baseline symmetry-aware architecture by replacing simple addition with conditional feature aggregation, which refines the HCOCF-enhanced features and improves boundary recovery. Extensive experiments on MSD, PMD, and RGBD-Mirror demonstrate competitive performance against representative supervised mirror segmentation methods. In particular, our RGB-only model achieves 88.47% IoU on MSD and 73.72% IoU on PMD, and remains competitive on RGBD-Mirror without using depth input. Full article

Soil release appearance grading evaluates residual stains after standardized laundering, but visual assessment is subjective and adjacent half grades are difficult to distinguish. This study proposes a lightweight Tri-Semantic Injection Network (TSI-Net) for nine-grade assessment. From one red–green–blue (RGB) image, a fixed CIE L*a*b* (CIELAB) branch constructs a mean-background image B, a pixel-wise color-difference image D, and a stain-appearance image S. The trainable backbone progressively injects S, D, and B and predicts grades from 1.0 to 5.0 at 0.5-grade intervals. Gaussian soft targets represent the ordering of these grades. The dataset contains 325 images, and TSI-Net has 1,403,336 trainable parameters. In a 33-image evaluation, Gaussian-trained TSI-Net achieved exact-grade and within-half-grade accuracies of 87.88% and 96.97%, compared with 84.85% and 93.94% for one-hot training. Both training methods achieved 100.00% accuracy within one grade. Gaussian training therefore showed a 3.03-percentage-point advantage for each of the two stricter metrics in this comparison. This method requires no manual stain segmentation and provides a compact framework for standardized fabric appearance assessment under controlled acquisition conditions. Full article

This study used 30 m Landsat imagery from 1995 to 2020 to investigate vegetation dynamics and their driving factors in Yengisar County, an arid oasis region in northwestern China. The Normalized Difference Vegetation Index (NDVI) was calculated, and the Dimidiate Pixel Model was used to estimate fractional vegetation cover (FVC). Trend analysis and a center-of-gravity migration model were employed to examine the spatiotemporal evolution of vegetation and its responses to climatic and human factors. The results showed that: (1) FVC exhibited clear spatial heterogeneity, with high vegetation coverage mainly distributed along river systems in the southwest and concentric expansion occurring around urban areas in the northeast. (2) From 1995 to 2020, the total vegetated area increased by 409.38 km², representing a 76.4% increase relative to 1995, while the average FVC increased by 15.4%. Except for the extremely low-coverage class, all vegetation classes expanded, with the medium-coverage class showing the largest increase (142.9%). (3) The vegetation center of gravity gradually shifted toward the southeast, and the period from 2015 to 2020 represented the most rapid stage of vegetation improvement. (4) Human activities, particularly agricultural reclamation and grassland expansion, exhibited a strong association with recent vegetation changes and played an important role in the expansion of the artificial oasis. The results indicate that recent oasis expansion in Yengisar County has been closely associated with land-use change and irrigation activities. These findings provide useful information for oasis management, land-use planning, and water resource allocation in arid regions. Full article

Reliable crop mapping is essential for understanding agricultural practices, optimizing resource use, and analyzing rural dynamics, while also supporting modelling and sustainable agriculture planning. However, obtaining 10 m crop distribution maps remains challenging in Mediterranean regions, where data availability is often limited and landscapes are fragmented. The main European land use dataset, Corine Land Cover (CLC), lacks both the crop specificity required for accurate crop differentiation and the temporal frequency needed for timely monitoring. This study addresses these limitations by implementing the IOTA² automated chain in Sardinia (Italy), to create a large-scale crop map specifically targeting Mediterranean crops. The methodology leverages open-source satellite imagery with supervised machine learning, using the 2018 Land Parcel Identification System (LPIS), CLC, and Urban Atlas dataset for training. We compared two nomenclatures, detailed (32 classes) versus simplified (25 classes), testing each across three training sample sizes (10%, 50%, and 100%). Results indicate that the simplified nomenclature (N25) provided more robust performances, achieving an overall accuracy (OA) of 0.77 with full sampling, compared to 0.61 for the detailed version. These OA values refer to the subset of reference polygons held out from the reference data for independent pixel-level validation. Moreover, a final map was produced using the entire reference dataset for training and evaluated through zonal area agreement. Mapping showed high performance for specific crops like rice, citrus, and grapevine, while classes such as cereals and fruit trees presented classification challenges due to high fragmentation of the landscape and irregular crop-distribution patterns. Despite these challenges, this work delivers a 10 m spatial resolution reproducible framework that enhances thematic details of current European datasets. By running as a single automated processing chain rather than a sequence of manually executed steps, it offers a scalable solution for rapid, annual crop monitoring in complex, data-scarce Mediterranean environments. Full article

Medical images such as Magnetic Resonance Imaging (MRI) scans consist of various confidential data that can hamper the privacy of the patient if leaked. Organizations have to abide by various international standards in order to ensure the safe storage and transmission of medical data. Strong encryption is a solution; however, conventional full-image encryption which treats every pixel as equally sensitive results in high computational costs. The resulting energy and carbon emissions pose an environmental challenge. Adaptive selective encryption secures images based on the sensitivity of different portions, resulting in lower carbon emissions. However, these claims have rarely been validated in existing studies covering both real adaptive detection performance and fair energy accounting. To alleviate these issues, this study presents DNA-AdCrypt, a carbon-aware deoxyribonucleic acid (DNA)-inspired tiered encryption system to secure MRI scans. The approach is simulated on brain tumors based on the Ultralytics MRI scan dataset. Fine-tuned You Only Look Once (YOLO) detectors are used to segment the scans into three domains, based on which the strength of the cryptosystem changes. YOLOv8n outperforms other variants, with the highest mAP50 (0.538) and mAP50-95 (0.396). In addition, DNA-Adcrypt reduces the encryption time by up to ≈49% and remains competitive in per-image energy against different simulated ciphers. This study puts forward a design intended to generalize for different modalities in order to promote a greener deployable medical security standard, with brain tumor MRI as an initial validated case. Full article

Existing architectural backdoors embed malicious logic directly into model structures to persist after clean training, but they typically rely on handcrafted trigger patterns vulnerable to preprocessing and lack the mechanism to dynamically update attack targets. Addressing these limitations, this paper proposes ArchLock, a dynamic-target architectural backdoor framework designed exclusively for stateful inference deployments that maintain persistent mutable state across requests. ArchLock comprises two core components: a correlation-based statistical trigger detector that utilizes local red-green channel Pearson correlations as the activation signal, thereby reducing dependence on absolute pixel values and exhibiting theoretical invariance to linear transformations; and an Adaptive Confidence Calibration module, which maintains a persistent memory buffer to enable post-deployment target switching through a two-phase protocol. Extensive experiments on CIFAR-10, CIFAR-100, Tiny-ImageNet, and ImageNet-100 demonstrate that ArchLock achieves a dynamic target success rate exceeding 90% while preserving clean accuracy. Furthermore, the method exhibits robustness against fine-tuning and pruning and evades three behavioral detectors under the stated stateful deployment assumptions. This work highlights the security risks of mutable architectural components in stateful inference services while explicitly discussing the limitations of the proposed approach under stateless deployment and aggressive preprocessing conditions. Full article

The 2021 eruption of Cumbre Vieja on La Palma, Canary Islands, produced extensive lava flows and major landscape transformation, providing an opportunity to evaluate remote sensing approaches for mapping the extent of an emplaced lava flow. This study assessed direct spectral, classic Rao’s quadratic entropy (RaoQ), and multidimensional RaoQ approaches using satellite observations acquired before and after the eruption. Optical, radar, thermal infrared, and night-time radiance datasets were evaluated within a common change-detection framework implemented in Google Earth Engine. Difference maps were converted into binary change maps using a histogram-based thresholding procedure calibrated on the reference delineation and evaluated against the Copernicus Emergency Management Service (CEMS) lava-flow reference and no-change validation areas derived from ESA WorldCover using multiple accuracy metrics. Because the change reference is the final CEMS lava-flow delineation and the no-change samples lie outside a 100 m buffer around it, the accuracy figures reported here quantify the mapping of lava-flow extent and not of other eruption-related effects such as ash deposition or vegetation damage beyond the flow margins. Among the direct spectral approaches, the NHI_SWIR index achieved the highest overall classification performance. Among the individual Sentinel-2 bands, B12 achieved the highest overall accuracy, whereas B8A achieved the highest true skill statistic; both exceeded the multidimensional RaoQ configurations in mean prevalence-independent discrimination. Within the classic RaoQ approach, MIRBI produced the strongest single-variable heterogeneity-based results. The best multidimensional configurations combined Sentinel-2 B8A and B12 with Sentinel-1 VV, demonstrating that radar backscatter provided complementary information to optical observations. Although multidimensional RaoQ did not surpass the best direct spectral variables, it produced competitive and spatially coherent representations of lava-flow disturbance. The evaluated thermal infrared and night-time radiance products did not provide competitive discrimination under the selected spatial and temporal conditions for different reasons: a thresholding limitation in the case of the Landsat thermal product, and an unfavourable ratio of pixel size to flow width in the case of the night-time radiance products, while the MODIS product returned no valid validation points and could not be evaluated. These product-specific explanations rest on a small number of comparisons and are provisional. These results show that carefully selected Sentinel-2 SWIR variables remain the strongest benchmark for detailed mapping of fresh lava-flow disturbance, while multidimensional RaoQ provides a framework for optical–radar integration that requires no training data or prior classification. Because the evaluation covers a single eruption in a single landscape, transfer of the framework to other events and settings remains to be demonstrated. Full article

Accurate stereo visual–inertial measurement is essential for mobile robots operating in Global Navigation Satellite System (GNSS)-denied and structurally complex environments. In stereo visual–inertial odometry (VIO), pose and trajectory outputs depend strongly on the point measurements delivered by the visual front end before sensor-fusion update. In sparse-texture but structurally regular scenes, tracked point features may exhibit poor persistence, uneven spatial distribution, and local tracking noise, even when informative line structures are present. Existing point–line VIO methods can improve positioning accuracy by introducing line landmarks or line residuals, but they usually modify the estimator state, measurement model, and Jacobian treatment. We present a scene-adaptive line-aware visual measurement conditioning method for stereo VIO front ends with point-measurement updates. The method uses 2-D image-line segments as lightweight structural priors and applies bounded normal-direction conditioning to reliable point measurements before a fixed-interface VIO back-end update. A sparse pruning safeguard removes only highly inconsistent long-lived tracks under strong structural support, while a scene-level confidence gate attenuates the intervention when line evidence is weak or unstable. The method is instantiated and evaluated in an S-MSCKF pipeline. On the reported EuRoC MAV sequences, it reduces the sequence-averaged absolute trajectory error (ATE) RMSE by approximately 13% relative to S-MSCKF, with 3–27% reductions on machine-hall sequences. On three real-world robot measurement sequences with an RTK-aided inertial reference, the mean Sim(2)-aligned planar position error decreases from 8.72 m to 7.31 m, and the mean yaw error decreases from $8.{02}^{\circ }$ to $6.{76}^{\circ }$; an additional scale-preserving SE(2) evaluation reveals sequence-dependent planar behavior and residual metric-scale sensitivity. Candidate-level stereo-consistency diagnostics show subpixel mean and 95th-percentile image-domain perturbations without systematic vertical-stereo bias, while the final reliability-weighted primary-view update is analytically bounded by approximately 0.221 pixels in the reported implementation. Runtime profiling reports an average front-end time of 33.34 ms on the tested CPU platform, close to the 33.3 ms frame period of the 30 Hz stereo input, although the $\mu +3\sigma$ runtime of 47.23 ms exceeds a strict frame-by-frame 30 Hz budget. These results suggest that line-aware front-end conditioning can improve visual measurement quality in structured stereo visual–inertial sensing without modifying the evaluated back-end interface. Full article

Accurate instance-level perception of aerial view traffic accident scenes is the foundation of accident investigation. However, existing accident datasets mainly support event-level video analysis or coarse spatial localization, and rarely provide pixel-level vehicle masks together with accident-involved labels. To address this gap, we construct the Drone-oriented Accident Recognition and Segmentation (DARS) dataset, an instance segmentation dataset specifically developed for aerial view traffic accident scenes. DARS contains 7603 images and 49,613 vehicle instances with six classes jointly defined by vehicle type and accident-involved status. Statistical analysis reveals the class distribution, scale variation, and accident-type composition of the dataset. We further introduce a Frequency-Guided Adaptive Downsampling (FGAD) method into YOLO26n-seg to improve hierarchical feature extraction. FGAD performs content-adaptive aggregation of spatial candidate features, while wavelet-derived frequency information guides candidate weight estimation and provides an additional residual pathway. On DARS, the proposed method achieves 52.37% recall, 51.24% mAP@50, 41.18% mAP@75, and 36.77% mAP@50–95, outperforming other methods in terms of instance segmentation accuracy. On a UAV-based Vehicle Segmentation Dataset (UVSD), it consistently improves over YOLO26n-seg, reaching 75.61%, 57.74%, and 44.60%, respectively. These results support DARS as an instance-level benchmark and demonstrate the applicability of FGAD to aerial traffic perception, providing a basis for fine-grained accident scene analysis in intelligent and sustainable transportation systems. Full article