2

Where Can I find algorithm details for holistic word recognition? I need to build a simple OCR system in hardware (FPGAs actually), and the scientific journals seems so abstract?

Are there any open source (open core) codes for holistic word recognition?

Thanks

toolic
  • 57,801
  • 17
  • 75
  • 117
The Byzantine
  • 619
  • 1
  • 6
  • 21
  • @Joe Blow: FPGAs are Field-Programmable Gate Arrays. And they are simply integrated circuits that can be configured by the user. – Lukasz Dec 22 '10 at 10:44
  • @Joe: A Field Programmable Gate Array is "programmable hardware" -- it's a device with a bunch of logic gates (AND, OR, etc.) that you can programmatically "wire up" however you want (a bit like uploading firmware), use, and then "rewire" for something else. – j_random_hacker Dec 22 '10 at 10:45
  • 1
    @Joe are you taking the proverbial? – William Dec 22 '10 at 22:15
  • @Joe: I don't know, but this would make a good SO question in its own right. – j_random_hacker Dec 29 '10 at 01:57

1 Answers1

4

For an algorithm that is quite suitable for FPGA implementation (embarrassingly parallel) you might look at:

http://en.wikipedia.org/wiki/Cross-correlation

It is fast, and easily implemented.

The only thing is: it recognizes a shape (in your case some text) DEPENDENT of the rotation and size / stretch / skew etc. But if that isn't a problem, it can be very fast and is quite robust. You should only watch out for interpretation problems with characters that are similar (like o and c).

I used it to find default texts on scanned forms to obtain bearings where Region of Interests are and searching in those images (6M pixels) only took around 15 ms with our implementation on a Core2 CPU in a single thread.

Ritsaert Hornstra
  • 5,013
  • 1
  • 33
  • 51
  • 1
    We implemented the formula from: http://scribblethink.org/Work/nvisionInterface/nip.html and used a mipmap / image pyramid like scheme to make sure that the number of FLOPs was kept low – Ritsaert Hornstra Dec 22 '10 at 21:51