How to get all tables in an ocr pdf as attached, and then parse all cells to get coordinates, then use the coordinates to re-ocr with another engine, then update the ocr result in pdf?

[nissansz]

table2TextOverImage.pdf

How to get all tables in an ocr pdf as attached, and then parse all cells to get coordinates, then use the coordinates to re-ocr with another engine, then update the ocr result in pdf?

[JorjMcKie]

Use tabs = page.find_tables(), then tab = tabs[0] to access the table itself.
The list tab.cells contains a rect-like tuple for each table cell.
I cannot answer the remaining part of the question.