The pictures are already inside the PDF you uploaded, so lifting them out costs nothing and calls no AI. What is left over is the only part anyone has to do by hand.
On the content model, find the Pictures panel at the top right.
It is there only when the book has pictures, and counts how many are placed.
Press Take the Pictures from the PDF.
Pick the right one where a page held several.
Thumbnails with their sizes. Choosing for you would sometimes put the wrong diagram in a lesson, so it asks.
Note anything on the by-hand list.
A diagram drawn as lines rather than saved as a picture has nothing to lift out. Those stay as marked placeholders and are replaced in the Lesson Editor after the write.
A picture used by two lessons is stored once - the library keeps one copy of an identical image however many lessons show it.
A very large picture is shrunk to fit. It is stored with the lesson, the same way a picture pasted into the Lesson Editor is.