Talk:MUTCD/California

Just thinking out loud... it would be an interesting project for some enterprising person to develop an app that allows you to take a picture of the sign and then provides the corresponding OSM tags. Maybe even integrated into GoMap! or Vespucci where after taking a picture you can click a button and enter a node of that sign. I know very little about image processing, but it seems like you could use ML to train on the known signs and since they are standardized and little generalization is required the accuracy would be high. But again, out of my league. BubbleGuppies (talk) 05:45, 7 May 2022 (UTC)