CoreDFX

Apply Intelligence – Core Analysis on Autopilot

Deploy trained AI models to extract geological insights directly from core imagery. Whether it's segmenting structural features, classifying rock types, or calculating fracture properties, MLange simplifies it to just a few clicks. Select from a library of ready-to-use workflows—each crafted by geoscientists for real-world exploration. Input your RGB, depth, or geometry images, and instantly receive classified outputs like rock type tables, geometry layers, or fracture orientation CSVs. This module empowers your team to run high-resolution interpretations consistently, repeatably, and without needing coding.

Click. Train. Predict – Build AI, Not Code

Design and train Machine/Deep Learning models tailored for drill core data with no coding required. This feature supports a wide range of tasks including image classification, object detection, semantic segmentation, instance segmentation, regression, and more. Easily import labeled images, select your modeling task, and choose from proven deep learning architectures like CNNs, U-Nets, and YOLO.

MLange streamlines the process of dataset handling, training, evaluation, and model export—making it simple for geoscientists, researchers, and technicians to build AI tools that accelerate core logging, alteration mapping, mineral detection, and other high-resolution analysis tasks.

Scan. Sense. Sync – Unified Core Data Capture

Collect high-resolution visual and sensor data from drill cores with unmatched precision and flexibility. MLange’s data collection module integrates multiple capture modes—from standard snapshots and macro imaging to flythroughs and overlapping scans. You can also pair imaging with advanced sensor modalities like magnetic susceptibility, infrared spectroscopy, LIBS/XRF, and UCS strength testing.

Whether you're mapping textures, identifying mineralogy, or assessing geomechanics, this tool ensures that every core is documented in rich, co-registered detail—ready for AI analysis or expert interpretation. Just choose your method, select your tray layout, and let MLange handle the rest.

Label Smart – Train Models Your Way

Create rich, AI-ready annotations with precision and ease. MLange’s annotation studio is purpose-built for geological imagery and supports pixel-level segmentation, object bounding boxes, polygon labeling, keypoints, and text regions. Whether you're preparing data for core feature segmentation, fracture detection, or OCR-based logging, MLange ensures your labels are interoperable with major deep learning ecosystems.

Seamlessly export annotations to formats like COCO JSON, YOLO TXT, Pascal VOC XML, EasyOCR-compatible folders, or even custom .npy and .coredfx files. Use annotation layers as ground truth for image classification, instance segmentation, regression, and more. Train once, reuse everywhere—MLange bridges expert annotation with modern AI pipeline.

GeoStats & Downhole Insights – Visualize the Invisible

Turn drillhole data into geological insight. This module empowers geoscientists to explore lithology, alteration, fractures, geochemistry, and mineralization patterns through synchronized downhole logs and interactive geostatistical plots.

MLange’s visual analytics bridge spatial data and geochemical signatures—revealing subtle correlations, zoning, and structural controls on mineralization. Load your .csv, .las, or .coredfx datasets and let the patterns surface naturally.

Database Builder – Structure Your Exploration Data with Ease

Design and manage your drill core database without writing a single line of SQL. MLange’s Database Builder empowers geologists and data managers to define table schemas, map CSVs to structured fields, and unify fragmented datasets into a clean, query-ready format. Set primary keys, data types, and units directly in the visual interface, while importing multiple CSVs using intuitive drag-and-drop mapping.

Whether you're organizing drillhole metadata, logging intervals, or planning analytical jobs, this tool ensures your database is robust, interoperable, and ready for downstream workflows like model training, visualization, and reporting.

Stay on Top of Every Drillhole

Plan, monitor, and manage your entire drill core workflow from acquisition to AI output. MLange’s CoreTrack Manager gives you complete control over logging jobs, box-level metadata, image acquisition programs (macro, snapshot, RGBD), and QA/QC checkpoints like geometric alignment and recovery validation. Easily view progress across depth intervals, drillholes, or trays, and access linked data products like feature maps, recovery estimates, and fracture analysis in one place.

Designed for geologists, supervisors, and data engineers, this module ensures your data is traceable, organized, and always audit-ready—from field to model deployment.

Report Viewer – Visualize, Verify, and Share Results Instantly

MLange’s Report Viewer brings together structured summaries, depth intervals, imagery, 3D models, and AI-driven outputs into beautifully formatted reports—ready for validation or stakeholder review. View per-box statistics such as RQD, core recovery, fracture count, and lithology in context with macro imagery, depth color maps, and fracture orientation reconstructions.

Generated as interactive HTML, each report links back to the original dataset, enabling rapid verification, decision-making, and collaboration between geologists, data scientists, and project leads. No formatting required—just generate, review, and share.

Spectral Viewer – Seamless Visualization of Hyperspectral Images

Unlock hidden mineralogical signatures with MLange’s Spectral Viewer. Load hyperspectral images and visualize them in customizable RGB composites or single-band enhancements. Instantly pair visual scans with precise infrared reflectance spectra to identify diagnostic absorption features and validate mineral zones in real time.

Compatible with ENVI .hdr formats and tailored for drill core trays. Whether you're identifying muscovite, chlorite, alunite, or detecting subtle shifts in alteration patterns, MLange provides the tools to see what your eyes can’t.

3D Core viewer

Experience your core data in 3D. The MLange 3D Viewer provides a spatially accurate reconstruction of core trays using RGB-D or photogrammetric input. Fracture planes (in red) are automatically detected and mapped, enabling geologists to assess orientation, continuity, and volume relationships intuitively.

Whether you're assessing core recovery or analyzing structural features, the MLange 3D Viewer bridges visual analysis with quantitative insight.

MLange