OCR on scanned images.

[Jaime Rodriguez]

Can I perform OCR on document images that have been scanned with a regular scanner? What are the requirements for the image to be analyzed using the Vision Assistant in the Vision Development Module for Labview?

This is for an application the involves offline inspection of labels for packages. I am wondering if I could just use a scanner instead of a camera?

[ryanm]

Hi Jaime,

Vision Assistant will allow you to load .tif, .bmp, .png, and .apd images.  Assuming your scanner can save images as one of these formats, you will be able to load the image into Vision Assistant.  In order to perform OCR, you will need a 8-bit or binary image.  If you have a color images, there are several options to convert to 8-bit or binary images including extracting a color plane and thresholding.  Once this processing has been performed, you will be able to run OCR.

Regards,

Ryan M.
Applications Engineer
National Instruments