OCR only recognizing some text and labeling others as images

[Bridget23453089l1ba]

I have a document created in Micorsoft Publisher that needs to be converted to an accessible PDF for website publishing.  When I've gone into Adobe Acrobat Pro DC and tried to run the accessibility tools, some pages don't allow me to select the text boxes on the page.  I've tried running in through the recognize text feature, and while some text boxes are recognized, others are only seen as images and therefore not able to edit.  And then a few pages just recognize the whole page as an image, with none of the individual elements showing up as specific boxes.  Is there any way to tell the program either in Acrobat or maybe in Publisher that they need to be kept as text boxes?

[Luke Jennings3]

Are you able to save or export the Publisher file as a pdf? That may preserve the live type. If you see an option to embed fonts, say yes, if your pdf has fonts from Publisher that are not embedded, you can (and should) possibly fix that in Acrobat with a preflight profile. Can you explain exactly how you are creating your pdf from Publisher? Perhaps someone will have some specific tips for you.