Extracting map data from images

You can use Mapillary to edit maps or geospatial datasets using street-level imagery as a visual guide for reference. Street-level imagery allows you to check ground conditions, providing context for what’s on the map. Furthermore, our technology automatically extracts map data from imagery so you can save up on manual labor.

You can always use your own imagery on Mapillary any way you want. You can also use public Mapillary imagery that comes with the CC-BY SA license. 

With Mapillary, you can get a look at not only the ground level but even the specific locations of your GIS point data in order to make sure it is accurately placed. You can also filter Mapillary imagery by date and username—ensuring that the images reflect the relevant timeframe for your data, and even getting a look at the images you personally uploaded.

This process is possible through the Mapillary platform as well as several tools. Popular options for this are Mapillary for ArcGIS Online and Mapillary for ArcGIS Pro, which enable you to load your own maps to view, edit, and create features. In addition, our imagery is available for OpenStreetMap projects using the iD editor or JOSM.

Mapillary also offers different types of map data that are automatically extracted from Mapillary imagery.

We use advanced computer vision technology to detect and extract valuable data from the contributed images on our platform. We provide object detections (the result of semantic segmentation) and map features in GIS-friendly formats (shapefiles, GeoJSON).

For more advanced methods of retrieving data using the Mapillary API, please see our developer resources.