Newish OSMapper here, but longtime contributor to Google Maps. I have a 360 cam used for bike trails and roads, plus have added many images of stores/amenities etc. I spend me time primarily in three locations and have been adding/updating those areas on OSM (building outlines, parks, roads, rivers).
I just learned of the existence of OSC and Mapillary and noticed that only one of my locations is well mapped in Mapillary, and none of them (except the major highways) have been mapped in OSC. Would someone please explain the use of (and maybe difference between) these platforms? I started using OSC to submit the roads in my hometown, but I don’t know if that’s meant for my own purposes or to allow other people to make edits from home. I can’t think of a reason I’d need to have a photo of my own city when I can just bike there to make the change, so I stopped submitting images. I’m not sure why to adopt either, because I don’t fully understand what is done with the end result.
Advise would be appreciated, and if this is not the best place to ask, I would be happy to be redirected to the appropriate community/forum.
Second, Open Street Cam (OSC) vs. Mapillary vs. Google Streetview… I can only answer this for me personally:
Google Streetview
Pro: smooth, very well programmed for all platforms
Con: Completely closed source, zero customization possible. Wanna use it, pay for it. Google… they altered their motto in 2015 from “Don’t be evil” to “Do the right thing”, they should have altered it to: “lett’s make the internet a Google app”…
I am a strong believer of diversity, in all things… also on the web. Google is way to big, way to powerful and in the long run will damage the whole “internet concept” (which in my opinion is now).
That is why I now first look elsewhere and when I find a competitive and/or promising product/service I tend to choose the later.
Open Street Cam (OSC)
Pro: truly opensource
Con: goal? Application poor, functionality is poor.
I don’t see a vision, why is OSC doing what it’s doing?. The application is poor. Tools for integrations are (extremely) limited. API seems primarily focused on collecting?
Mapillary
Pro: relatively smooth, works great on some platforms, a bit less on others. Semi open source. An API with nice publishing capabilities and great potential. There is a good concept / business plan that pays for the platform (long term viability).
Con: Semi open source. buggy on some browsers/platforms. Sometimes you feel “the work in progress”.
Up until now I don’t see any problem with the “Semi” part in “Semi open source”. Based upon their manifesto I expect it’ll keep that way. The website app of Mapilarry works very nicely. Since there aren’t many API implementations “in the wild” I think there hasn’t been much “vibe” for that part (yet)
I especially like the API, I’ve got plans with that
So that is why I choose to invest time, equipement, energy in Mapillary and not in one of the others
google: take users work, make money from it, do not publish und a open source license.
aka: We take the free work of users and sell it back to them with them not being allowed to use it for free
Open Street Cam - a consortium from car makers - the goal is simple: take pictures from the users for better self driving cars. But using opensource license, aka users may use the pictures free of charge for own project.
One point I saw: snatch to road is a pity. E.g. I made a track from railroad, and it was mapped automaticly to streets “nearby”. completely worthless
mapillary - company taking pictures from users to create 3d map with enriched metadata for others to buy. Nearly same like OSC, but independent from car manufactors. Also opensource.
So, in the end, it does not matter, do not use google, use OSC and/or mapillary. If you like extra work, you can upload your pictures to both of them.
Mapillary is a computer vision company, so our focus is to scale the collection of map data by recognising map features in street-level images. We believe that the collaborative model is the best way to do this. Part of this approach means making map data available to OpenStreetMap. Traffic signs are already available in OpenStreetMap, and we’re also running experiments to see how we can make either map features that we detect automatically like bicycle racks, mailboxes, and benches available within OpenStreetMap.
Why is it worth collecting street-level imagery?
You can collect a lot more OSM data than you would manually.
It’s available for others so that they can also map the area.
Mapillary identified objects automatically which you can then view and add to the map.
You have a historical record of how an area changes over time. You can use Mapillary’s Time Travel feature to compare.
there might be things in your photos that you wouldnt think of adding that others might. like somebody might want to know whether the crosswalks have tactile pavement, somebody else might want to add cycle lanes, bus stop numbers or the position of electricity poles. when you think about each person having to go around their town/city each time they want to add certain things, its much more efficient to have 1 person take photos and then have other people reference them
even just for myself, there were always things i would forget to check when i was in town, like did a bus stop have a bench or a bin etc. and there would be other things i would see on my way that i would end up forgetting by the time i got home so thats mainly why i started using mapillary
the OSC map is empty where i live as well. so was mapillary when i started but there seemed to be more of a community around mapillary, it had a more advanced app and website, and i liked that is was more of a neutral company compared to OSC… which is why i chose mapillary in the end
I consider OSC to be more roadway (and traffic sign detection) oriented while Mapillary is a bit more broad. Overall, they are pretty similar from an OSM perspective. They are both meant for your own use and others - I frequently rely upon other user’s images while editing OSM.
I can’t think of a reason I’d need to have a photo of my own city when I can just bike there to make the change, so I stopped submitting images. I’m not sure why to adopt either, because I don’t fully understand what is done with the end result.
As another user mentioned, your focus may be different than other users. Even if you have a great memory and spend extensive time mapping and area, uploading an image allows others to contribute too. You may be interested in mapping roads and trees but someone else may be wondering what the roadway surface material is, or differences between the seasons.
Advise would be appreciated, and if this is not the best place to ask, I would be happy to be redirected to the appropriate community/forum.
This is a fine place to ask. For OSM-specific discussions (and OSC) the OSM-US Slack group is a great resource: https://slack.openstreetmap.us/
