
Scaling map data generation using computer vision
The MapScaping Podcast - GIS, Geospatial, Remote Sensing, earth observation and digital geography
00:00
How to Collect Data for Post Processing
Using our apps is by far the easiest way to collect. We build a 3D model of every location where someone takes pictures. Our current production system does 97 different unique classes. So we have 1500 unique signs in 100 countries that are being automatically indexed and labeled. The other things, if we see them in more than one image, we can also position and give them their own latitude and longitude. And then they become map features.
Play episode from 10:56
Transcript


