Receipt scanner + parser
Allows to upload an image file and then choose settings for text extraction method.
The result is displayed in a redactable table, that can be saved as a .csv
(or anything) later.
Future work: allow to change GUI's language.
preproc
files do the preprocessing steps such that crop the image of a receipt and turn it into a scan-like image.
Before and after:
ocr
does the text extraction.
There are several methods:
just pass the scan
result to ocr
and let it to its job;
pass lines with text line by line and extract text from each of them separately;
pass lines (which are just a set of words) and extract text word by word.
parser
saves the extracted data into a .csv
file in the format Item,Price
.
Future improvements are to be made.
opencv tesseract leptonica dear imgui (opengl + glwf3) win32