-3

I am working on an application where I need to get the net price displayed in any shopping bill from its picture. I have already retrieved the editable text from the bill images using "tesseract ocr" API. Now I need to print only the "grand total amount" from the text. How do I extract only that part( total price) from a whole bill having the item name, quantity and price?

Eric Aya
  • 69,473
  • 35
  • 181
  • 253
Krishna Jayachandran
  • 25
  • 5
  • Find the text "grand total amount" and show the number next to it? – Bryan Chen Dec 19 '16 at 03:22
  • I don't know what the Tesseract API is capable of, but just some thoughts: (1) it's usually the highest number on the bill, (2) it's usually printed in big or bold fonts, (3) it usually appears at the top or bottom of the paper – Code Different Dec 19 '16 at 03:45
  • Thank you! I will try these out. Can you suggest me some algorithm that will work right for bills of different types ? – Krishna Jayachandran Dec 19 '16 at 05:21

1 Answers1

0

Short answer, I don't think there is a quick/handy method you can call directly.

You need to look into the .hocr file returned from Tesseract(You can google hocr for more info first). The .hocr includes all the bounding box of the text(x, y, width, height, language etc.) then make use of these values, you can determine if words are on the same line (The word 'Total' and the total amount are very likely printed on the same line).

From here you can shortlist the words, add some logical operations (maybe remove all characters/words), then you can get the total value.

ps: My company is working on a similar stuff, but we decided not to use Tesseract, as it is kind of slow and not easy to train (we're dealing with receipts in several languages). We are using Google Vision API.

Hope my answer helps :D

Pang Ho Ming
  • 1,299
  • 10
  • 29