Fine-grained 3D phenotypic analysis of rice plays a vital role in rice breeding and yield estimation. However, a comprehensive rice data acquisition and segmentation pipeline is still lacking. While Neural Radiance Fields (NeRF) have shown impressive results in crop-level 3D reconstruction, their high sensitivity to data volume and camera viewpoints often leads to reconstruction failures for rice. In addition, the large-scale rice point clouds, coupled with heavy occlusion and visual similarity among grains, pose significant challenges for fine-grained trait extraction. To address the challenge of reconstructing rice point clouds under low-quality data conditions, we propose a novel method named Multi-Scale NeRF(MSNeRF). This method incorporates a structure-detail collaborative reconstruction mechanism and a dynamic initialization density scheduling strategy. Furthermore, we introduce a multimodal and multitask rice dataset (MMR) as a benchmark resource for future research. For rice point cloud segmentation, we develop Vision Rice Knowledge Graph Network(VRKGNet), which comprises an image segmentation module, a projection module, and a point cloud segmentation module enhanced with a Transformer to enlarge the receptive field. VRKGNet performs standalone point cloud segmentation and integrates image segmentation results from multiple viewpoints as prior knowledge to enhance semantic and instance-level segmentation. Extensive experiments demonstrate that MSNeRF achieves high-fidelity point cloud reconstruction with as few as 10 viewpoints. VRKGNet achieves superior rice plant segmentation with a semantic segmentation mIoU of 88.79% and an instance segmentation AP 25 of 84.55%, outperforming mainstream algorithms.

Through the perspective of LiDAR: A feature-enriched and uncertainty-aware annotation pipeline for terrestrial point cloud segmentation

Despite significant advancements, segmentation based on deep neural networks in medical and surgical imaging faces several challenges, two of which we aim to address in this work. First, acquiring complete pixel-level segmentation labels for medical images is time-consuming and requires domain expertise. Second, typical segmentation pipelines cannot detect out-of-distribution (OOD) pixels, leaving them prone to spurious outputs during deployment. In this work, we propose a novel segmentation approach which broadly falls within the positive-unlabelled (PU) learning paradigm and exploits tools from OOD detection techniques. Our framework learns only from sparsely annotated pixels from multiple positive-only classes and does not use any annotation for the background class. These multi-class positive annotations naturally fall within the in-distribution (ID) set. Unlabelled pixels may contain positive classes but also negative ones, including what is typically referred to as background in standard segmentation formulations. To the best of our knowledge, this work is the first to formulate multi-class segmentation with sparse positive-only annotations as a pixel-wise PU learning problem and to address it using OOD detection techniques. Here, we forgo the need for background annotation and consider these together with any other unseen classes as part of the OOD set. Our framework can integrate, at a pixel-level, any OOD detection approaches designed for classification tasks. To address the lack of existing OOD datasets and established evaluation metric for medical image segmentation, we propose a cross-validation strategy that treats held-out labelled classes as OOD. Extensive experiments on both multi-class hyperspectral and RGB surgical imaging datasets demonstrate the robustness and generalisation capability of our proposed framework.

This article presents a Prairie-wide spatial vector dataset of agricultural field boundaries across Alberta, Saskatchewan, and Manitoba, Canada. The dataset was generated from Sentinel-2 Level-2A surface reflectance imagery (10 m spatial resolution) using an automated segmentation workflow based on the Segment Anything Model version 2 (SAM2). Sentinel-2 imagery was accessed in Google Earth Engine (GEE), filtered using cloud/quality masks, and aggregated into seasonal RGB composites representing key crop phenological periods (early-, mid-, and late-season) for large-scale segmentation input. The composite images were exported and processed with a SAM2 segmentation pipeline (tiled inference and mosaic-based post-processing) to delineate candidate field units without manually labeled training samples. Segmentation outputs were then post-processed using rule-based filtering and topology repair (removal of small artifacts/sliver polygons, hole filling, boundary cleaning, and geometry validity correction). Final vector outputs are distributed in ESRI Shapefile and GeoParquet formats with geometry attributes for downstream spatial analysis. The workflow code and processing scripts are provided to support reproducibility and adaptation to other regions. This dataset provides a consistent field-scale boundary reference layer for agricultural monitoring, crop and yield modeling, soil and environmental analysis, cropland mapping, land management, and machine-learning applications across the Canadian Prairies. •Prairie-wide field boundary datasets generated using automated segmentation of seasonal Sentinel-2 composites with the Segment Anything Model version 2, followed by vectorization and topological post-processing.•Vector outputs provided in both Shapefile and GeoParquet formats, enabling efficient use in traditional GIS, cloud-native, and big-data geospatial workflows.•A consistent, large-scale field boundary reference dataset supporting agricultural analysis, spatial modeling, cropland mapping, and machine learning applications across the Canadian Prairies.

Sensitivity analysis of a patient-specific finite element simulation pipeline incorporating automated vertebral segmentation for predicting vertebral strength of a clinical database.

We are developing a decision support system for treatment response assessment of bladder cancer by analyzing patients' CT urography (CTU) examinations. Accurate segmentation of bladder lesions is a critical and challenging task. We previously developed a bladder cancer segmentation method using a deep learning convolutional neural network and level sets (DL-CNN + LS). In this study, we designed several deep learning models based on U-Net for bladder cancer segmentation and compared them with DL-CNN + LS and two transformer-based models developed for medical imaging - DATTNet and the Med-Segment Anything Model (Med-SAM). Our new U-Net models did not use the second-stage level set refinement, greatly simplifying the overall segmentation pipeline. We trained and evaluated the models by using radiologist's hand-drawn 3D contours as the reference standard. The proposed Crop U-Net model, utilizing a user-defined box to direct the U-Net attention to the lesion region by masking out the structured background, was superior to other models being investigated. On the independent test set, the Crop U-Net achieved average Jaccard index (AJI) of 48.1 ± 18.0% and average minimum distance (AMD) of 4.3 ± 3.0mm, while the DL-CNN + LS achieved AJI of 33.2 ± 20.0% and AMD of 5.3 ± 2.2mm. The results demonstrated that the Crop U-Net could achieve a higher accuracy than the previous DL-CNN + LS while reducing the complexity of the segmentation pipeline.

Patent text segmentation is a fundamental task in patent data mining, enabling applications such as patent analysis and search. The objective is to decompose structurally complex, lengthy sentences into grammatically complete, semantically equivalent short sentences to facilitate downstream processing. Traditional approaches rely on manually defined rules or feature-based machine learning methods, which are labor-intensive, domain-specific, and exhibit limited generalizability. To overcome these limitations, this study proposes a Deep Segmentation Model for Patent Text (DS2PT), a two-stage fine-grained segmentation framework based on ALBERT. The first stage employs a conditional random field model to perform coarse segmentation of patent paragraphs into shorter clauses based on structural cues. The second stage utilizes the ALBERT model to perform deep, context-aware segmentation of complex clauses into syntactically independent and semantically complete sentences. Compared to conventional methods, DS2PT effectively captures hierarchical contextual information across two stages, significantly improving segmentation accuracy without semantic loss. Furthermore, this research draws inspiration from advancements in cross-lingual speech-to-text systems with low-latency neural networks for real-time applications. While the domains differ, the core technical challenges are analogous: both require models to process sequential, information-dense input (audio streams or long sentences) into structured, meaningful units (transcribed text or segmented clauses) with high accuracy and efficiency. The principles of low-latency neural networks-such as efficient context modeling, parallelizable architectures, and real-time incremental processing-inform the design of our segmentation pipeline to enhance its scalability and potential for integration into real-time patent analysis systems. Similarly, the cross-lingual capability highlights the importance of model generalization, which aligns with our goal of developing a domain-adaptive segmentation tool for diverse patent corpora.

The anterior approach (AA) with liver hanging maneuver (LHM) has been proposed as an alternative to the conventional approach (CA) for major hepatectomies. Despite its potential advantages, LHM remains underutilized, partly due to concerns about vascular injury and tumor rupture. Three-dimensional visualization technology (3DVT) may improve anatomical comprehension and inform preoperative decision-making in selecting surgical strategies. We retrospectively analyzed 20 patients undergoing major hepatic resections (right/left hepatectomy and right posterior sectionectomy) between 2019 and 2024. Four expert hepatobiliary surgeons(HPB) and four postgraduate surgical trainees (PGY5) independently assessed surgical strategy based on 2D imaging, followed by reevaluation with 3D reconstructions. Patient-specific 3D structures were generated using an AI-assisted segmentation pipeline and systematically revised by physician specialists, a board-certified abdominal radiologist and two hepatobiliary surgeons. Intra-rater concordance was evaluated using Cohen's Kappa. Primary endpoint was the rate and directionality of surgical plan modifications due to 3DVT. Secondary endpoints included perioperative outcomes and segmentation performance metrics. 3DVT prompted significant changes in surgical planning, particularly in assessing LHM feasibility. Trainees exhibited a higher proportion of positive shifts in decision-making (No → Yes: 17.6%) compared to experts (10.8%), whereas experts more frequently reversed previously affirmative decisions (Yes → No: 9.5%). In select raters, negative Kappa values indicated systematic reassessment driven by 3D data. No significant differences in intraoperative blood loss, operative time, transfusion rate, complications, or mortality were observed between AA + LHM and CA cohorts. 3D segmentation achieved high concordance with manual ground truth (median Dice similarity coefficient for liver parenchyma: 0.98). 3DVT exerts a quantifiable influence on preoperative strategy, particularly for complex hepatic resections. It facilitates surgical planning among trainees and enhances precision among experienced surgeons. Integration of 3DVT may support safer adoption of technically demanding maneuvers such as LHM, especially in minimally invasive settings.

Quantitative RNA imaging in large plant tissues has historically been challenging because of limited spatial resolution, low signal-to-noise ratios, and the largely qualitative nature of traditional RNA in situ hybridization methods. Hybridization chain reaction-RNA fluorescence in situ hybridization (HCR RNA-FISH) coupled with high-resolution microscopy enables sensitive detection of RNA molecules at cellular resolution. However, quantitative approaches that combine improved tissue accessibility with robust computational pipelines for single-cell transcript quantification in plants remain limited. Here, we present a quantitative HCR RNA-FISH protocol for the developing maize inflorescence. We describe a 4-day workflow that includes fixation, agarose immobilization, vibratome sectioning, probe hybridization, amplification, and mounting and enables multiplexed detection of transcripts at the cellular level in maize ear and tassel primordia. In addition, we provide a Python-based image analysis pipeline for (i) cell segmentation, (ii) RNA spot quantification, (iii) assignment of spots to cells, and (iv) data representation. The scripts can be easily run on Jupyter notebooks and are available on GitHub. Overall, this protocol highlights the importance of integrating robust imaging strategies with quantitative and reproducible data analysis frameworks to extract biologically meaningful insights from imaging data. © 2026 Wiley Periodicals LLC. Basic Protocol: Quantitative RNA imaging in sections of maize ear and tassel primordia using HCR RNA-FISH.

In this investigation, a Deep Learning (DL) approach was applied to measure the morphometry of Transverse Aeolian Ridges (TARs) on the surface of Mars. A large sample of TARs was segmented from High Resolution Imaging Science Experiment (HiRISE) images, the highest resolution remote sensing dataset presently available for the planet. HiRISE images located between 50°N and 50°S, and from all longitudes were selected. Morphometric parameters such as area, elongation, and orientation were retrieved for this sample using a supervised instance segmentation and geospatial analysis pipeline. The result is the most extensive catalogue of TAR morphometry to date extracted from ~14 million candidate TARs in ~7000 HiRISE Images. This was accomplished by training off-the-shelf DL models within a Geographic Information System (GIS) environment. A significant TAR population was found in approximately half of the images surveyed. TAR area, and the lengths of the long and short axes, were found to exhibit a positively skewed log-normal distribution; the median short axis length is 5 m, while the median long axis is 24 m. Median elongation is 0.24. Global TAR orientations are varied, although North-South oriented TARs are the most populous group. This is likely due to the strong east blowing winds predicted by GCM simulations of the modern martian climate. Here we present our latest results and use TAR orientation statistics to describe the emerging picture of global wind patterns on Mars during TAR forming epochs. • TARs have been segmented with a Mask R-CNN model, using ArcGIS Pro. • 14 million TARs were measured in ~7000 HiRISE images of the mid latitudes of Mars. • Orientation and morphometry data were retrieved using an automated DL/GIS pipeline. • TAR orientation provides a geomorphic markers for past wind conditions on Mars. • Comparisons to manual digitisations show the results to be representative and fit for purpose.

Deep Learning-Driven Transmission Electron Microscopy Analysis of Murine Optic Nerve Myelinated Axons.

Efficient phenotyping is essential for accelerating genetic improvement in turfgrass breeding, where manual measurements are labor-intensive. This study evaluated hyperspectral imaging (HSI) as a high-throughput tool for assessing Zoysia spp. breeding populations consisting of 464 genotypes. HSI data (400–1000 nm) were processed through a user-in-the-loop hybrid segmentation pipeline integrating UMAP dimensionality reduction, DBSCAN clustering, Random Forest classification, and pseudo-RGB refinement. To independently assess vegetation classification performance, 10,000 manually annotated reference points from 50 pseudo-RGB images were compared with the automated module, yielding an overall accuracy of 0.9697, a precision of 0.8830, a recall of 0.9240, a specificity of 0.9779, an F1-score of 0.9030, and Cohen’s kappa of 0.8851. A Combined Ranking Score (CRS) integrating five vegetation indices and vegetation pixel count was significantly associated with aerial shoot count (r = −0.445, p < 0.001) and runner count (r = −0.207, p < 0.001). The highest-ranked genotype showed a 9370.3-pixel increase in vegetation area between 6 and 16 weeks after transplanting, compared with 1417.7 pixels for the lowest-ranked genotype. Classification performance declined under high-coverage conditions, indicating increased mixed-pixel ambiguity in dense canopies. These results suggest that HSI-based CRS can support rapid, objective, and non-destructive relative ranking of density-related vegetative growth in turfgrass breeding. Because the study was conducted at a single location and season and correlations with manual traits were moderate, the framework is best interpreted as a screening and ranking tool rather than a direct predictive model.

• Non-destructive microfluidic imaging for pore-scale tracking of microplastics. • Integrated visualization–quantification framework links pore morphology to MP fate. • Heterogeneous pore networks act as long-term sinks for microplastics. • Connectivity-driven networks act as high-mobility conduits for microplastics. • Uniform throat networks act as reversible reservoirs of microplastic retention. Microplastics are persistent contaminants in soils and aquifers, accumulating and potentially remobilizing within groundwater and surface flows. Predicting their fate requires pore-scale insights in realistic porous media, especially under saturation-desaturation conditions. Despite the growing body of research on MP transport in porous media, the methodological divide between qualitative visualization and quantitative measurement limit predictive understanding under realistic subsurface conditions. Microfluidic systems enable direct pore-scale observation of MP mobility but rarely provide domain-wide quantification, while column-scale experiments offer robust mass balance yet obscure the underlying pore-scale mechanisms. This disconnect hinders the development of mechanistic understanding linking pore geometry, multiphase flow, and MPs fate. To address this limitation, this study introduces an integrated, non-destructive microfluidic imaging framework that couples high-resolution time-lapse visualization with semi-automated image analysis to simultaneously resolve pore-scale processes and quantify domain-scale retention across saturation-desaturation cycle. Four statistically distinct micromodels representing heterogeneous porous domains were used to capture realistic flow and retention behaviors. The segmentation pipeline classified individual phases (air, water, solid, and MPs), allowing spatially and temporally resolved quantification of retention and release throughout the entire flow field. The results demonstrate that MPs mobility depend not on porosity alone but on the combined effects of throat-size distribution, connectivity, and capillary dynamics, which control the persistence or remobilization of particles under transient saturation. Beyond its immediate findings, this work establishes a reproducible and scalable experimental-computational framework that bridges the gap between visualization and quantification, offering a mechanistic basis for predicting MPs fate in soils and aquifers under dynamic environmental conditions.

Neuroimaging datasets are increasingly shared in open repositories for research purposes, raising concerns about participant re-identification through facial features visible in brain magnetic resonance imaging (MRI) scans. MRI defacing algorithms address this risk by obscuring identifiable facial structures while preserving brain tissue for analysis. Algorithm performance varies substantially by context. For re-identification prevention, fsl_deface and mri_reface achieve the lowest recognition rates, while afni_refacer and pydeface demonstrate the highest processing success rates. However, all algorithms affect brain volumetric measurements to varying degrees, with some causing failures in automated segmentation pipelines. Performance is notably age-dependent, with specific algorithms underperforming in pediatric or elderly cohorts and in clinical populations with neurological disorders. Optimal algorithm selection depends on research priorities. For preserving brain measurements, mri_reface and SPM-based defacing are preferred; for pediatric studies, FreeSurfer better preserves brain voxels; for electroencephalography (EEG) and magnetoencephalography (MEG) co-registration, AnonyMI provides superior geometrical preservation. This review examines the major defacing algorithms and their validation across diverse datasets, evaluating effectiveness in preventing re-identification, preserving brain measurements, and maintaining compatibility across age groups. A comparative discussion highlights the trade-offs between privacy protection and data utility, emphasizing the need for a study-specific approach when selecting a defacing method.

To quantitatively evaluate global and regional brain volumes in children with biochemically confirmed growth hormone deficiency (GHD) using standardized volumetricMRI. This retrospective case-control study included 64 children with isolated GHD (43 boys, 21 girls; mean age, 11.4±3.2 years) and 64 age- and sex-matched healthy controls. High-resolution 3D T1-weighted MRI scans were processed using the volBrain automated segmentation pipeline. Global, cortical, and subcortical volumes were normalized to intracranial cavity volume (ICV). Group differences were analyzed with age-, sex-, and ICV-adjusted models, corrected for multiple comparisons (FDR<0.05). Children with GHD exhibited significantly smaller gray matter (750.6±92.3 vs. 789.0±88.5 cm3; q=0.002) and total brain (1,204.3±124.9 vs. 1,259.2±118.6 cm3; q=0.003) volumes compared with controls, while white matter volume was preserved. Regionally, hippocampal (q=0.02), thalamic (q=0.03), and caudate (q=0.04) volumes were smaller in GHD. ICV-normalized models showed reduced gray matter and limbic shares, with relatively higher thalamic and white matter proportions. Differences were more pronounced in prepubertal subjects and partially attenuated during puberty. Pediatric GHD is associated with selective reductions in gray matter-dominant and limbic-striatal regions, with relative preservation of white matter. Quantitative volumetric MRI provides a reproducible biomarker framework for assessing GH-related neurodevelopmental alterations.

Ischemic stroke represents a leading cause of global mortality and long-term disability. Consequently, rapid quantification of cerebral damage severity, coupled with timely neuromodulatory intervention, is imperative for improving clinical prognosis. In this study, we present a label-free quantitative monitoring framework utilizing optical-resolution photoacoustic microscopy (OR-PAM) to assess microvascular alterations and evaluate the therapeutic efficacy of low-intensity transcranial ultrasound stimulation (LITUS). A graded ischemic stroke model was established by modulating photothrombotic irradiation duration (3, 5, and 10 min) to validate system sensitivity. Subsequently, a Hessian filter-based segmentation pipeline was employed to extract quantitative vascular metrics. Five key morphological and functional parameters, including the photoacoustic (PA) signal intensity, vessel area fraction (VAF), average vessel diameter (AVD), branch-point number (BN), and perfused vessel density (PVD), were extracted to characterize the severity of ischemic injury. Our analysis revealed a strong negative correlation between irradiation duration and vascular perfusion-related metrics, which was further confirmed by ex vivo 2,3,5-triphenyltetrazolium chloride (TTC) staining showing duration-dependent infarct expansion. Applying this quantitative framework to the established model, we further demonstrated that acute LITUS treatment significantly alleviated microvascular hypoperfusion and promoted rapid hemodynamic recovery by restoring both functional perfusion and vascular morphology via vasodilation. These findings highlight the potential of a PAM-based quantitative framework for accurate grading of ischemic severity and evaluation of ultrasound-based neuromodulation.

Deploying conditional Generative Adversarial Networks (cGANs) for industrial texture synthesis faces two barriers: the prohibitive cost of manual data annotation and the uncertain alignment between automated evaluation metrics and human perception. This study addresses both challenges for marble texture synthesis using 289 high-resolution industrial scans. We adapt an unsupervised segmentation pipeline combining Simple Linear Iterative Clustering (SLIC) superpixels, Gaussian Mixture Models (GMMs), and graph cut optimization to extract vein structures without manual annotation. Four cGAN architectures—baseline cGAN, Pix2Pix, BicycleGAN, and GauGAN—are benchmarked using a dual-evaluation protocol contrasting ten automated metrics with structured human-centered assessment. The results reveal a significant metric–perception discrepancy. Pix2Pix achieved the best Fréchet Inception Distance (FID = 85.3) yet received the lowest human ratings due to periodic texture artifacts. GauGAN produced textures statistically indistinguishable from real marble, achieving a Visual Turing Pass Rate (VTPR) of 0.533 and a Mean Opinion Score on Marble Authenticity (MOS-MA) of 2.89, despite an inferior FID (87.3). These findings make three contributions: an annotation-free segmentation pipeline, empirical evidence that automated metrics alone are insufficient for architecture selection, and a dual-evaluation framework that establishes human-in-the-loop assessment as essential for quality-critical industrial deployment.

Mitochondrial networks exhibit striking heterogeneity in their morphology and distribution across different neuronal compartments, reflecting the diverse metabolic demands of these structures. In this study, we used automated tape-collecting ultramicrotome scanning electron microscopy (ATUM-SEM) to reconstruct and quantify mitochondrial networks in the somata and neurites of neurons in the rat prefrontal cortex (PFC) and hippocampus (HPC; CA1 stratum radiatum). We developed an automated segmentation pipeline based on an attention-enhanced 3D U-Net to extract all mitochondria from volumetric EM data. Our quantitative analyses revealed pronounced regional and subcellular heterogeneity. In the PFC, the mitochondrial volume fraction was higher in neurites (7.2%) than in somata (2.9%; 7.1% when nucleus was excluded). Mean individual mitochondrial volume was 0.11 μm³ for neuritic and 0.33 μm³ for somatic mitochondria in the PFC, with similar results observed in the HPC (0.13 μm³ in neurites, 0.31 μm³ in somata). In both regions, the vast majority of mitochondria (~91%) assumed an oval or rod shape, with few displaying branched or donut-shaped structures (~1%). Notably, elongated linear mitochondria (~8%) were mostly confined to neurites, and approximately 90% of these comprised up to 120 nanotunnels-thin segments (<220 nm) connecting enlarged, oval-shaped structures (>350 nm) in tandem. These data provide a detailed quantitative characterization of mitochondrial network architecture in the adult rat cortex and hippocampus, revealing significant regional and subcellular differences in mitochondrial morphology and distribution.

Under growing urbanization and environmental challenges, sustainable urban mobility has become a critical priority for cities worldwide. Public Transport (PT) systems play a central role in reducing car dependency, lowering emissions, increasing network capacity, and promoting more equitable and efficient access to urban spaces for all users. Hence, the present paper aims to investigate PT preferences in the city of Larissa, Greece. Larissa is a medium-sized city currently serviced only by buses, and is currently focusing on the potential introduction of a new tram system to operate in parallel with existing bus services. To this end, a SP survey was designed and implemented, resulting in 972 observations that were collected for further statistical analysis. Survey results show a slight preference for trams over buses, with 54.63% selecting the tram and 45.37% favoring the buses. Moreover, a context-based segmentation pipeline was established using PCA, DBSCAN and t-SNE algorithms, aiding the visualization of existing clusters for transport choice approaches. Afterwards, a series of mixed logit models was applied, and statistically significant variables influencing mode choice were determined. The study also examines Value of Time (VoT) metrics and finds that respondents assign lower VoTs to trams than to buses, especially in out-of-vehicle segments of the journey, such as waiting and walking, and therefore consider trams as more pleasant and less burdensome. The findings also indicate that passengers place a high value on the quality of infrastructure related to access and waiting times, underlining the need to improve the overall user experience beyond the vehicle itself. In summary, the present research offers valuable insights into how the introduction of a tram system could possibly reshape PT usage patterns when compared with the legacy existing bus services.

ABSTRACT Extensive resin wood formation in Scots pine stems infected by Cronartium pini constitutes a significant quality defect that compromises timber value and complicates industrial processing decisions. While X-ray computed tomography (CT) offers non-destructive density mapping capabilities, the wood processing industry currently lacks validated, automated segmentation methods to differentiate resin wood from unaffected heartwood and sapwood. A hybrid automated segmentation pipeline incorporating an adaptive per-slice Gaussian mixture model (GMM) was developed to combine statistical parameterisation of tissue-specific intensity distributions with multi-directional ray-casting for structural likelihood scoring. The framework utilises unsupervised image processing methodologies and morphological refinement to improve boundary delineation between complex internal wood features. Validation against manual annotations by trained technicians demonstrated high precision (0.94) for sapwood segmentation and substantial inter-rater agreement (Cohen’s κ = 0.819). However, precision for resin wood was limited to 0.27, reflecting the inherent difficulty of distinguishing pathological resin saturation from natural impregnation of heartwood. By implementing heatmap-based logic to overcome the limitations of simple intensity thresholding, this research establishes a foundational open-source pathway for automated internal resin wood mapping. The proposed pipeline offers a computationally efficient approach for research-scale analysis and provides a viable strategy towards potential value recovery in sawmills.