Can Textract extract Maps and images from PDFs?

1. unfortunately, there is no recommendation. As an alternative question, I recommend describing what you would like to achieve and solicit ideas.

2. textract does indeed recognize tables. It finds what it considers to be a table, predicts what the structure will be, and returns a tree structure. It does not extract the table as is.

https://docs.aws.amazon.com/textract/latest/dg/how-it-works-tables.html