Geospatial Business Spotlight: Mapillary

Company Name:                 Mapillary

Website:                              http://www.mapillary.com/

Established:                       2013

THE COMPANY

Mapillary is a new and different approach to Street View. Using computer vision, Mapillary stitches together photos taken with any device to create street-level imagery for extracting geospatial data. By empowering anyone anywhere to easily create street-level imagery, Mapillary aims to create a photo representation of the world. To date, Mapillary’s community has contributed 63 million photos spanning over 900,000 miles across all seven continents.

Mapillary was founded in 2013 by Jan Erik Solem, Johan Gyllenspetz, Yubin Kuang, and Peter Neubauer, who share a vision for putting mapmaking into the hands of people everywhere. Solem, who serves as CEO, previously founded Polar Rose, a facial recognition software that was bought by Apple in 2010.  The company is headquartered in Malmö, Sweden, and has fifteen full-time employees located around the world.   Mapillary’s New York office is based In Brooklyn, NY.

PRODUCTS AND SERVICES

Mapillary allows community members to create and explore a crowd-sourced, street-level view of the world through its app, which is available for web, iOS, Android, and Windows Phone. Users can upload photos from their smartphones or other digital cameras and Mapillary then combines the images together geotagging each using GPS metadata. The resulting 3D reconstruction is created using computer vision and the company’s own open source “Structure from Motion” algorithm. Mapping isn’t limited to streets—the community has captured hiking trails, favorite bike routes, and even a stretch of Antarctica. Mapillary integrates with any mapping platform through simple APIs.

Street-level photos can be captured using Mapillary's iPhone app

Street-level photos can be captured using Mapillary’s iPhone app

In addition to creating photo-maps, Mapillary also utilizes computer vision techniques to extract useful data from uploaded images. For instance, the software can recognize symbols on street signs in photos taken in the U.S. or Europe. This capability grew out of the app’s need to blur faces and license plates for privacy purposes, and is now used by groups like city governments for urban planning, land surveys and asset inventory. Using Mapillary with Esri’s ArcGIS platform has helped cities streamline infrastructure updates, from speed limit changes to road surface quality checks.

Mapillary uses computer vision to automatically detect traffic signs and extract geospatial data

Mapillary uses computer vision to automatically detect traffic signs and extract geospatial data

Humanitarian organizations also use Mapillary to further their initiatives. In Haiti, the American Red Cross and Humanitarian OpenStreetMap partnered with Mapillary to help record previously unmapped areas to aid disaster response efforts. Mapillary has also been used by the World Bank and other groups in Dar Es Salaam, Tanzania, to record infrastructure in flood-prone areas of the city to help plan improvements. The app is ideally suited for these sorts of tasks because it doesn’t require any special equipment beyond a basic smartphone.

Street level imagery is reconstructed by creating point clouds from photos.

Street level imagery is reconstructed by creating point clouds from photos.

Mapillary is free to use for individuals and for non-profit or educational purposes. For commercial solutions, ranging from embedding a photo-map into a webpage to extracting specific data from images, Mapillary has several pricing plans available.

Westchester County GIS is using the Mapillary app to inventory hiking trails in County parks

Westchester County GIS is using the Mapillary app to inventory  hiking trails in Westchester County parks

To find out more about Mapillary, visit their website.

CONTACT: 

hello@mapillary.com
Bredgatan 4, 211 30 Malmo, Sweden
www.mapillary.com

Leave a Reply

Your email address will not be published. Required fields are marked *

Follow Me
Follow Me