DOCX processing doesn't set pages on the doclingdocument

### Question
I am trying to get images for all the pages in my document, the process them, to determine bounding boxes for headers and footers.

My end goal is, to remove all repeated, items like logos, headers, page counter etc. As the documents I try to convert have different layouts, I need to determine the bounding boxes for each of them individually. Hence I try to process the page image to get the bounding boxes.

I am using the `generate_page_images=True` attribute on the pipeline options. For PDF documents it does indeed return the pages as images (DoclingDocument.document.pages). However running the processing for docx documents, the mentioned array is empty. No page images are returned.

Could you please point out if I am missing something, or is it just how the system works ?