The detailed technical description of is project is:
PDF --> text --> text-to-speech --> audio --> saved audio file.
There is existing software for each stage that I would integrate. Adobe
has an Acrobat SDK, which can extract text from a PDF. The text can be
converted to speech with the Microsoft text-to-speech software or maybe
even Jaws. The generated speech can be played immediately or saved to file
(WAV or Mp3 format). The different components can be combined using Microsoft
tools, such as Visual Basic or Visual Studio C++.
PDF 2 text
An Adobe SDK provides a programmable interface to the Adobe Acrobat
Reader and Abode Acrobat (the full version). The SDK should provide access
to the text of the PDF. There are other solutions on the web to access
text from pdf. Here are some solutions.
Text-to-speech
Microsoft provides free text-to-speech software. I imagine possible
enhancements to simply reading the text with one voice. The voice could
change from a man's voice to a woman's voice when reading titles, italics,
or other outstanding text.
Save speech to audio file
The MS text-to-speech generates the audio and can also store the audio
to a WAV file. I am sure code exists to convert the WAV to mp3. Once in
Mp3, I could imagine downloading the audio to an MP3 player and listening
to the PDF paper while walking outside, riding the bus, or in the car.