tablecv 0.1.0

Table extraction from image.

TableCV

Extract table from an image.

Usage

There are two ways to get table from an image.

Approach 1 (uses PaddleOCR)

Call extract_table (returns pandas Dataframe object).

from tablecv import extract_table

print(extract_table(image_path=""))

Approach 2

Perform ocr using your favourite OCR tool (EasyOCR, KerasOCR, PaddleOCR, WhateverOCR ...).

ocr_results object should look like the following:

# list of tuple of bounding box and text

ocr_results = [
    (
        (1, 2, 3, 4), "a"   # (x, y, w, h), text
    ),
    (
        (4, 5, 6, 7), "b"
    ),
    ...
]

and then call extract_table_from_ocr method.

from tablecv import extract_table_from_ocr

ocr_results: list[tuple[tuple[float, float, float, float], str]] = ...
print(extract_table_from_ocr(ocr_results))