Support additional basemap layers

[matthiasschaub]

We currently only support two basemap layers that tend not be very accessible for users are for instance visually impaired or have not acquired map literacy yet. This excludes / makes it harder to participate for certain groups to work with the generated sketchmaps.

The heal client has implemented some potential basemaps.
https://heal.openrouteservice.org/#/place/@8.68112,49.410757,12

Maybe as well https://opentopomap.org/#map=5/49.023/10.020

Related: #363

---

This will not be worked on until requested by uses (more requirements have been collected)

[lysogeny]

First of all, this is a very cute tool you folks have made and I quite like it!

Having just seen someone use the sketchmap tool, it strikes me as a bit odd that both of the map layers seem to be fairly opaque and quite colourful. From my anecdotal observation this causes strokes that are drawn on the map with a pen to be difficult to distinguish by humans and the underlying computer vision models used in segmentation also seem to struggle with that on certain colours a bit. This seems to particularly be the case for colours that are commonly line-shaped in the base map, like orange and light blue.

I would think that maybe the CartoDB positron basemap might provide really good contrast for pens and possibly make it easier for the computer vision models to pick up the pen strokes. For satellite imagery, offering desaturated versions might also help.

While I am fairly confident that some less opaque base layers might make the overall sketches more human-friendly, I am now also quite curious if it would improve the underlying computer vision model's ability to distinguish penstrokes.

[Chwiggy]

Hi I was recently digitizing maps on which students sketched positive and negative bikeability and I was wondering how the carto db osm base-map layer influenced both their recollection as well as their drawing, and how different their drawings might have looked if there was a bike focused base layer.

For instance, just as important roads and intersections for cars are very prominent on that base layer, the students identified a lot of troublesome intersections for their biking experience, but maybe on a base layer with more prominent bike paths maybe they would have identified other pain points more readily.

In the end, this is, i think, a more general phenomenon. OSM carto cannot be a neutral objective representation of the world as it is (nor is satelite imagery) but it's inevitably coloured (literally) by specific interests, decisions and priorities, that might not match with the perspective of users. And while that will be true for all base maps, having a selection of different base maps might at least give users some more agency to choose which base layer matches their particular interest