How Accurate is OCR?
There are many factors that contribute to the success of an OCR project, and all of them inevitably affect the cost of the service.

• The most important factor is the print quality of the documents that you send us. For example, a page that is output from a 600-dpi laser printer on white paper will have nearly perfect recognition compared to a fax-quality image containing blurry or shadowy text.

• Assuming good print quality, the second major factor is the nature of the text. For instance, regular text (e.g., "The quick brown fox jumped over the lazy dog.") is ideal, and results in perfect recognition. However, text containing special characters, bullet symbols, non-English words, chemical formulas, math symbols, subscript/superscript characters, etc. are much more difficult to OCR and require much more manual labor.

• The final factors we consider are page layout and binding. The golden rule is: Simpler is Better. A single-column text-only page is much easier to process than a multi-column catalog page that contains text wrapped around images. Likewise, single-sided loose leaf pages are easier to process than double-sided bound volumes.

This is not to say that some pages cannot be OCRd and converted. It simply means that the process can range from simple to very difficult.

What Does It Cost?
We are not able to firm-quote any OCR project until we see the entire document, however, we can usually provide an estimate based on a few sample pages from the document. In most cases, we will ask you to fax over a few "representative" pages. The base cost for an OCR job can range from $0.30/pg to several dollars per page depending on the nature of the document and the final file format desired (i.e. pdf, .doc, etc).

"I Have Ten Zillion Pages to Scan..."
RSA is not a large-volume OCR shop. Our services are aimed at small companies and colleges who have jobs no larger than a few thousand pages. Please contact us to make sure.