DigitalGlobe and Ecopia Produce US Building Footprints Dataset with Machine Learning - 15/10/2018


DigitalGlobe has announced that they have produced the first precision, GIS-ready building footprints dataset of the entire United States in partnership with Ecopia Tech Corporation.

Ecopia U.S. Building Footprints, available for order, includes more than 169 million building footprints created by proprietary artificial intelligence algorithms in combination with DigitalGlobe's high-resolution, high-accuracy satellite imagery. The detailed dataset helps customers save time and money by knowing exactly what the built environment looks like in their area of interest, allowing them to make decisions with confidence.

The 2D vector polygon dataset has a greater than 95% accuracy rating and will be refreshed every six months based on new DigitalGlobe imagery received through the company's Geospatial Big Data platform (GBDX), ensuring the most current foundational information is available to inform customer decisions on projects like mapping, insurance risk assessment, disaster management and land administration.

“The combination of Ecopia algorithms and DigitalGlobe's high-resolution satellite imagery accessed through GBDX creates reliable solutions for many industries. The Building Footprints puts readily available data in customers' hands to make timely business decisions and eliminates the need for manual, slow and expensive on-the-ground surveys.” said Bill Singleton, Ecopia Tech Corporation Vice President. “We are proud to bring a product to market that has the potential to transform many applications, such as reducing the time between census data collection from once per decade to once per year for civil governments.”

DigitalGlobe and Ecopia will extract building footprints for major international locations by the beginning of 2019. Additionally, the partnership is ready to produce building footprints anywhere in the world based on customer request. To learn more, visit www.digitalglobe.com/products/building-footprints

Last updated: 22/10/2018