Processing images about pdf2docx

[dothinking]

How exactly are the floating images?

It's a floating image If it overlaps with other elements like text or other images. Since they can't be arranged in flow mode -> one of them is floating. For instance, a background image behind some texts.

ValueError: no valid image found

So this has nothing to do with floating image since no overlaps exist in Page 1. It seems an upstream issue -> pdf2docx relies on PyMuPDF, a powerful PDF processing library, to extract texts and images. I raised an issue to PyMuPDF. Hope you don't mind since I used the image in test.pdf for sample.

pymupdf/PyMuPDF#699

[Jalkhov]

I understand, well, let's hope they take the problem into consideration, thanks for being attentive to the doubts and problems.