[Blindmath] Reading technical e-books with tables, charts, diagrams, etc

[Ben Humphreys]

Good morning listers,

I'm curious to know what you all do when you're happily plodding 
through an electronic book and you get to an inaccessible table, 
formula, graph, code sample, figure etc?

For the books I read, this happens about every 10 minutes.

So far, my strategy has been to buy books in Kindle format, remove 
the DRM, and extract the HTML and graphics.  Then read the text in browser.

The text in the browser typically indicates the graphic it refers to, 
so I'll go find the graphic and run it through Abbey OCR, with 
generally good results.

But this has several disadvantages:
* Momentum reading the text is lost while fighting graphics
* Text tables must still be read into Excel so one can navigate 
easily with spoken row and column headings
* OCR works nicely for tables, code samples but not for charts or diagrams
* OCR accuracy ranges from 0% to 97% but even if it's very close, a 
single 1 that turns into an i is problematic without human review

For charts and diagrams, I can envision someone skilled in Photoshop 
removing noisy backgrounds, enhancing important lines, and labeling 
important points in Braille font, then embossing.

I guess in a K-12 setting, a Teacher of Visually Impaired person 
would birddog students books and do these kind of things.

But once out of high school, where does one find such a person?

Thanks,

Ben