How to split a table into cells from a scan?

Question

Python

I wrote a python script that splits the scanned document into blocks and cuts them into separate files, these files are fed into Tesseract. The problem is the appearance of a table in the document, how can you understand that this is a table and cut it into cells (take into account collspan / rowspan)?

Answer 1

Pavel, where is "The problem is the appearance of the table in the document" is this?
Your code produces nparray - this is a set of numbers, but not table data.