Skip to content Skip to main navigation Report an accessibility issue

Open Source Streetview Imagery

One of the best parts about working for the GIS Outreach and Community Engagement Lab is having the opportunity to meet so many people who use geospatial technology in ways we’ve never heard of. At GIS Day in 2016 we learned about a company called Mapillary that makes it possible to collect your own open source street view style imagery. Since that time we’ve been testing out their mobile application to see if we could incorporate their tools into our education and outreach efforts. We were very fortunate to raise enough money in our Volstarter campaign to purchase the necessary equipment to collect 360o imagery, not only from cars, but also from bicycles and walking too. We plan on developing our protocols and procedures here on campus while collecting data for UTK Facilities Services and then expanding to community partners.

What exactly is going on in the image above? We were testing the mobile application for Mapillary here on campus so spent some time walking around with our phones collecting imagery. Once you collect the imagery it gets uploaded to Mapillary’s servers where they run a number of processes on the imagery, including trying to use machine learning to identify features in the image like crosswalks, sidewalks, vegetation, trees, and cars.