How can we make images with (necessary) text more translatable mechanically?
Our documentation set includes some diagrams, such as entity relationship diagrams and flowcharts, where text is integral and cannot reasonably be handled in callouts. Our documentation is translated into another language, so these diagrams ought to get translated too. (I don't know if they currently are.)
What can we do to make that translation easier from a technical perspective? In particular, we use Visio to create these diagrams. Is there a way that somebody who's not also using Visio could take our source files and replace the text strings? That is, could somebody do a text-replace on the strings in the file even without using the tool to edit graphically? I'm thinking that this might be easier to automate/repeat, plus I don't know what graphics tools the translators have.
In this related question I asked about supporting the needs of differently-verbose languages. In this question I'm asking about mechanics, not content.