Alternative Input Methods (OCR, Voice, ....)

[simonpoole]

This is a placeholder really so that it is clear that work is being done on the topic and we are not tied to classical touch screen keyboard input. Naturally lots of ideas sound good but fail in real life practical use.

http://master.apis.dev.openstreetmap.org/node/4299549413

http://master.apis.dev.openstreetmap.org/node/4299549410

Exploring these requires adressing some issues that are already on our todo list in particular https://github.com/MarcusWolschon/osmeditor4android/issues/316, https://github.com/MarcusWolschon/osmeditor4android/issues/223 and https://github.com/MarcusWolschon/osmeditor4android/issues/228

Some avenues being explored

• OCR: it is already possible to use an external OCR app by copying the recognized text and pasting from the system clipboard (0.9.6) in the property editor (which has always worked for individual text fields). It would be very easy to integrate this better, the main issue is that OCRing street signs gives very mixed results and is currently probably more work than simply typing. If somebody knows of an OCR app that works reliably for signs etc please come forward.
• Voice commands and inputs: work on a PoC is under way (and works for some limited use cases).
• Super simple touch screen/android wear interface: some of the same concepts as for voice input could be used for a three to four button touch screen interface, potentially combined with voice input.

I'll be updating this issue if and when things change.

[simonpoole]

https://github.com/MarcusWolschon/osmeditor4android/commit/64d865a19224bd97c80b845a8c28964c88ec4737 adds voice input support for creating nodes in the "New mode", you can now long press and then say "McDonald's" to get a certain burger restaurant :-)