extract_words() finding unexisting "blocks"

[Renato-Waldem]

Im having trouble while performing simple extract_words(). It is finding these "blocks" that doesnt have any character int it. Is there any paramenter to avoid those non-characters objects?

Thanks in advance

[jsvine]

Are you able to share the PDF itself, and the code that you're using?

One possibility is that the PDF does contain text there, but that it's just not visible. If you run select-all (something like Command-a, depending on the PDF viewer and operating system) on your PDF in a standard PDF viewer and then paste it into a text editor, do you see any text that might correspond to those areas?