Is there any way the computer vision tools can read bubbles?

Question

I have a string of computer vision tools to input pdf files and enhance them for OCR Optimization which seems to be performing decently against text. However, it seems not to recognize the big black dot next to some of the answer options on the form.

Has anyone encountered this issue?

Example here.

caltang · Accepted Answer

The black bubble isn’t a text, maybe that’s why it is not picking it up.

Side note, may I ask why you are reading a form’s response? Do you have access to the form results - because reading it via OCR seems extra.