Hello everyone,

I was wondering if anyone had any advice on using OCR to extract the
names of features on maps, so that we could automatically create a
searchable dataset with co-ordinates.

Has anyone heard of doing it the other way round?  That is to say,
instead of using OCR to get the names of features (which I'm guessing is
fairly hit and miss), you acquire an independent dataset of features
with longitude and latitude co-ordinates, and use that as your
searchable set, matching the co-ordinates in the dataset with those on
the map?

Any advice would be appreciated.

Kind regards,

Michael Paige,
Information Management Assistant,
Enterprise Information Unit,
Ministry of Finance

Frost Bldg North, 1st Floor
95 Grosvenor Street
Toronto ON M7A 1Z1

Tel: (416) 326 9443