The GEOINT Glossary
GSD
Ground Sample Distance: The distance between pixel centers measured on the ground. MetaVI is optimized for high performance even at challenging GSD (30-70cm).
Object Detection and Classification
The ability of AI to automatically identify objects of interest in aerial and satellite imagery, locate them within the image, and categorize them into mission-relevant classes. MetaVI’s no-code platform enables analysts to rapidly create and refine detection models without requiring data science expertise.
Imagery Interpretation
The process of analyzing aerial or satellite imagery to understand what is visible, identify relevant objects or patterns, and extract meaningful operational insights.
Geospatial Intelligence / GEOINT
Intelligence derived from imagery, geographic data, and location-based information. GEOINT supports decision-making by connecting visual information with its geographic context.
Analyst-in-the-Loop
An AI workflow in which the analyst remains central to the process. The analyst defines the mission, provides examples, reviews results, corrects errors, and improves the model over time.
No-Code AI
An AI workflow that enables users to build, train, evaluate, and deploy models without writing code or requiring deep AI expertise.
AI Model
A computational model trained to recognize patterns in data. In imagery analysis, an AI model can detect objects, classify areas, identify changes, or support interpretation tasks.
Model Training
Annotation
One-Click Labeling
Object Detection
Area of Interest / AOI
Actionable Insights
Deployable Model
On-Premises Deployment
Cloud Deployment
Air-Gapped Environment
Feedback Loop
Full AI Life Cycle
Insights & Case Studies
Metavi FAQ
Security & Infrastructure
Yes. MetaVI can be deployed in secure on-premises and air-gapped environments, meaning the platform can operate inside the customer’s protected infrastructure without requiring internet or cloud connectivity.
For customers who prefer a cloud-based setup, MetaVI can also be deployed in a secure cloud environment .
No. In on-premises and air-gapped deployments, MetaVI is designed so that imagery, annotations, training data, models, outputs remain within the customer’s security perimeter. The platform does not require customer data to be sent to MetaVI or to external cloud services.
For customers who choose MetaVI’s secure cloud deployment, the required data is uploaded to MetaVI’s secure cloud environment and managed according to the agreed security, privacy, compliance.
Technology, Performance & Integration
MetaVI is designed specifically for aerial and satellite imagery, where new mission requirements often appear before large labeled datasets are available. Its architecture allows the system to adapt quickly to a new object type from a small number of analyst-provided examples.
Instead of training a generic model from scratch, MetaVI uses task-specific adaptation, and domain-optimized model configuration to focus learning on the most informative examples. This enables the platform to detect relevant objects with high recall even when only limited labeled data is available.
This capability is the result of years of focused research and development, many accumulated person-years of engineering effort, and deep domain expertise in geospatial AI and operational imagery analysis.
MetaVI supports a wide range of aerial and satellite imagery sources, from high-resolution aerial collections to challenging satellite GSDs.
The platform integrates easily with standard geospatial environments, including OGC-compliant services, WMTS, COG, ArcGIS-based services, and customer-specific proprietary imagery streams.
Supported GSD depends on object size and visibility: if an analyst can reliably identify and label the object, MetaVI can train and optimize a model for that sensor, format, and resolution.
MetaVI is designed to integrate easily with the customer’s existing geospatial infrastructure, whether the imagery and data are delivered through standard geospatial formats and services or through customer-specific proprietary systems and streams.
This allows organizations to add AI-assisted imagery analysis on top of their existing GIS, C2, and analyst workflows, without replacing the systems they already use.