DigitalGlobe MDA Radiant Solutions SSL

Automated Feature Extraction & Object Detection

Cutting-edge technology that gets the most out of geospatial data at global scale

Computer Vision That Sees All.

Whether you are interested in monitoring discrete locations or looking for objects big and small, Radiant Solutions harnesses advancements in machine learning by applying computer vision and pattern analysis technology to geospatial problems throughout the entire data lifecycle.

Develop, deploy and validate custom algorithms and training data for feature extraction and object classification and labeling

Validate and optimize data via expert analysts and crowdsourcing supported by ground truth

2,349,695 features validated, 635,239 features invalidated and 525,419 new features identified in our first 627 crowdsourcing campaigns alone

Solutions:

Machine Learning

From producing labeled training datasets to developing, deploying and validating custom algorithms, Radiant Solutions delivers the technological and mission expertise needed to leverage machine learning and automation for game-changing results.

CoastLine

A revolutionary cloud-based system for mapping the world’s coastlines that connects users to open-source cloud services including third-party data, 24/7 satellite imagery feeds and automated coastline detection algorithms.

VRX

Video Reconnaissance Exploitation extracts value from videos and images faster than humans—enabling users to focus on the most relevant information to their mission. VRX’s detections can be tailored for customer needs, and can be visualized or interacted within various tools to support mission planning and operations, including Virtual Reality.

GeoHIVE Crowdsourcing

We’re experts at using public and private crowds to scale geospatial production tasks. Our patented Geospatial Human Imagery Verification Effort (GeoHIVE) brings together geospatial analysts, geospatial developers and imagery analysts to interact with a vetted crowd of online users to validate, discover or annotate features of interest in satellite imagery.