Blogs

OCR Stands for Automation – Not Literally

By Joe Budelli posted 09-29-2010 12:35

  

Two weeks ago, Michael Wells wrote a post for the Capture Community about removing the human.  I found it interesting for a number of reasons.  Not just because I am fond of robots, but because I appreciate the value of automated process in replacing the individual with more consistent and efficient alternatives, like computers.

The post also was of particular interest to me because of Michael’s inclusion of OCR as a good method for automating Metadata capture.  Obviously OCR is near and dear to my heart, as I am sure it is for all you reading. However, I feel that even though Michael included it in his list, the benefits of OCR were overshadowed.

Michael noted that OCR only works for certain types of documents, namely forms. Now he is correct that OCR only works for certain types of documents, however, the realm extends far beyond forms to include both structured, semi structured and un-structured document types.  This can include forms (invoices, purchase orders, work orders, etc.) but also includes books, reports, manuals, brochures and petitions, just to name a few.  Also, many OCR solutions also include ICR and fax recognition capabilities.  This means that the user can gain far more information and value from a document by moving beyond computer printed content. Are OCR solutions perfect? The easy answer is no, but they do deliver immense value previously unavailable to organizations.

Michael also raised the issue of scan operators needing to validate captures to ensure data integrity.  This is true.  However, Michael raises an interesting counterpoint earlier in his post with the fact that human data entry becomes laborious and typically leads to increased errors as the data entry carries on.   With human data entry, you not only need an operator to validate data, but you are also paying the individuals that manually enter the error-riddled data to begin with.  This can be avoided and simplified with users now having the ability to select the data they want to extract via easy to use touchscreens on their scanner – just like you would on an iPhone.

What are your thoughts? I see OCR solutions reaching success rates of 99.8 percent these days.  Doesn’t OCR seem like a great way to “remove the human” while driving down costs and increasing data capture accuracy?



#datacapture #ScanningandCapture #mobility #Scanners #documentcapture #OCR
0 comments
7 views