Footpath.ai imagery for OSM

Hi everyone,

I just wanted to introduce footpath.ai to the OSM community here. We collect pedestrian-focused street-level imagery with an open license for OSM. While we haven’t yet linked our images with any OSM editors automatically, you can view the imagery on our map browser to edit OSM data, particularly walking-related data, as our images are all captured on the footpath by walking mappers. You can find more about our work on our OSM Wiki page.

Our coverage is limited but growing. We have a good portion of Sydney, Melbourne, and Brisbane CBDs covered, with all images published on our map browser. Images from Adelaide and Perth CBDs will also be published soon. In Sydney, we also have the footpaths around a couple of suburbs outside the CBD mapped, with their imagery already published online. We actually have a lot more images in the queue to be processed and published.

We have been bulk uploading some of our data, including benches, bike racks, garbage bins, and trees in Sydney on OSM. Our most recent uploads on OSM included 1,732 trees and 316 benches in the Bondi and Waverley areas after data conflation and removing duplicates. We automatically detect street furniture and footpaths using computer vision, but before uploading to OSM, we perform human validation. However, we may still make mistakes with the OSM tags or false positives. We would appreciate the community’s review of our data to ensure errors are minimized. Any questions, comments, or feedback are always welcome.

We also love to connect more with the OSM community in Australia and New Zealand. With your help, we can expand the coverage much more. We work very similarly to Mapillary’s model, engaging with locals to collect imagery using action cameras like Insta360 or GoPro. The difference is we only focus on footpaths, and our images are captured by walking mappers. We have a growing walking community of mappers that you can sign up for if interested.

Cheers,
Meead

7 Likes