Identify any image.

A machine-learning image identification solution that accurately identifies image content.


A growing wealth of satellite and other overhead imagery is available for government and business intelligence. However, most of it goes untapped because of the labor-intensive task of labeling the content of the images. Organizations need innovative tools and capabilities to more effectively mine their imagery for mission-critical insights. Pixelate addresses these challenges using the latest advances in machine-learning and computer vision to identify, on its own, the content of images, rather than relying on analysts to do most of the work manually.

  • Accurate and automatic. Machine learning capability allows Pixelate to grow in accuracy and horsepower, enabling organizations to search vast amounts of imagery.
  • Enables analysts to focus on higher level search for insights. With the automated image search for identification and organization, analysts turn their attention to more strategic objectives.
  • Application across a wide range of clients and problems, from detecting the buildup of military assets along key borders to monitoring foreign naval base activity to identifying underground oil and gas reserves and measuring changes in supply of those same reserves.


Phi (Vu) Tran
Lead Scientist
To view all resources, please log in.


  • Deep machine learning auto-annotates images, improving accuracy and efficiency.
  • Annotation metadata enables search/retrieval, basic analysis and advanced change detection and alerting analytics.
  • Additional web applications are in development to identify types of objects in an image and support functions.