Zone OCR, or Optical Character Recognition, is a process by which a specific area on a scanned or imported document image is converted to text. Zoning, as it is called, is leveraged to automate data entry by mapping areas on a page to index fields. Straight zone OCR can be problematic due to a number of issues inherent to capturing documents:
- Page Registration – page registration is the shifting of document page images due to slight differences in scanning hardware. A minor shift in the image left or right can render zone OCR useless.
- Image Quality – low image quality can provide inaccurate results. Shading, speckling, contrast and brightness are just a few of the image problems that can affect the conversion of image to text.
- Form Changes – in forms processing, Zone OCR usually leverages a fixed image template for areas on a page. Any revision or change in the layout of the form can make zone OCR ineffective.
There are several enhanced zone OCR features that can alleviate these issues, including: image cleanup or processing, page anchoring and zone pattern matching.