0

I'm working on what is essentially scantron software. I'm writing in in python, and I've been trying to get optical recognition working. I've gotten to the point where given an image with a white background and a pencil mark on it, I can tell whether or not it is true/false. The next step is actually being able to read the form, and get each of the locations that need to be checked. I have full control over the design of the form, so now I need to find a way to operate relative to a sub-image.

Basically: given an image, I would like to be able to split up the image relative to a sub-image that I can search for. That way, I can have the forms use some sort of indicator at the top left (so that any minor changes in printing + scanning don't screw stuff up). Then, I need to be able to 'cut' the image relative to that sub-image. I know PIL allows you to chop up images, but I couldn't find a fuzzy subimage search feature... Any ideas on where I should go next?

I'm also open to other ideas on how to implement this.

arcologies
  • 742
  • 1
  • 6
  • 22
  • I don't understand your question much. But you can see a simple implementation of OCR in OpenCV-Python at http://stackoverflow.com/questions/9413216/simple-digit-recognition-ocr-in-opencv-python – Abid Rahman K May 04 '12 at 02:39
  • 1
    Check out [OpenCV](http://opencv.willowgarage.com/wiki/), it's a computer vision library with strong python support that has functions that'll let you search your image for known features (your registration mark). – Drakekin May 03 '12 at 07:38

0 Answers0