Jump to content

text recognition

- - - - -

This topic has been archived. This means that you cannot reply to this topic.
2 replies to this topic

#1
sebaka

sebaka

    Newbie

  • Members
  • Pip
  • 1 posts
hey guys...

i have to develop an application that will use scanned bills in order to extract the information in them (sums, total payments...) an insert them into a file (maybe excel) - because it takes a lot of time to do this by hand, an it's also error prone. I was wondering, whether this is feasable... Also, would you sugest writing the entire application, or using a OCR scanner that would provide the text recognition means...

thanks...

#2
WingedPanther

WingedPanther

    A spammer's worst nightmare

  • Moderators
  • 16,831 posts
Is it feasible? Yes. Is it even close to easy? No.

An OCR scanner will help a lot, but if the bills aren't in a consistent format, you're going to have a tough time getting the data correctly.
Programming is a branch of mathematics.
My CodeCall Blog | My Personal Blog

#3
Sinipull

Sinipull

    Programming Expert

  • Members
  • PipPipPipPipPipPip
  • 386 posts
Developing one right now (in Java), that can understand separate handwritten letters. It analyzes pictures and learns letter's structure and remembers it. It's much harder than i thought in the beginning, as program must recognize and make difference between horizontal, vertical and diagonal lines, which vary in thickness in different points. Currently it reads every pixel as part of the letter, which is not white (255,255,255).