Hello everyone,
How can I improve accuracy of tesseract while detecting text from retail receipts(walmart , Target etc) ?
How influential the fonts of receipt are in this matter?
I've developed an android application for the same, but it's not really successful when it comes to the mentioned receipts. It does better with office, home depot receipts. If it's related to fonts (assumption) how can I get around?