Whenever I use PDF-Tools (v6) to do OCR on a pdf, any images in that document get duplicated. This seems to make a trivial difference in terms of the size of the pdf, but it makes an annoying difference when it comes to text-to-audio software. It'll read each page that has image-text on it TWICE (it essentially reads the foreground image object, then the background image object after).
At first glance, I didn't even notice that it was duplicating the images, since the duplicated image object is layered exactly above the first. I was able to figure it out in editing though. I can fix the problem by just deleting the foreground image objects. Fortunately, the OCR ability is still retained after deleting the foreground image objects, and it stops reading pages twice afterward. However, it's a lot of extra work.
Why does this happen??? Can it be prevented???