Related Sponsors

OCR

Home » Industrial Equipment » Industrial Automation » OCR

OCR, or Optical Character Recognition, is the oldest form of data entry. OCR has been entering data in commercial and government EDP installations for years. OCR is becoming more and more popular with the advent of microprocessors, which allow for faster recognition techniques. Today’s OCR is reading print that at one time would have been rejected as to large. The rates of effectiveness and accuracy have also improved tremendously and OCR is now even more efficient than keystroke data entry. In fact, desktop OCR scanners can read typewritten data into a computer at rates up to 2400 words a minute.

OCR works through two basic methods. The first is Matrix matching. This is the simpler and more common form. With matrix matching, the OCR scanner compares what it reads to a library of character templates. When the computer finds a match within a given level of similarity, the image is labeled as the corresponding ASCII character. The second method is Feature Extraction. It may also be referred to as Intelligent Character Recognition (ICR) or Topical Feature Analysis. Instead of comparing what is read to a prescribed set of templates, the computer looks for general features such as open areas, closed shapes, or diagonal lines, and then reads the data based on how much “computer intelligence” the manufacturer has applied. Each method has its strengths. Matrix Matching is best when there is a limited repertoire of type styles. Topical Feature Analysis is superior for more unpredictable characters.

OCR offers a number of advantages over other forms of data entry. Using OCR reduces data entry errors, consolidates data entry, handles peak loads more efficiently, is human readable, can be used with many printing techniques, and is great for scanning corrections.

Next Page: RFID

Related Industrial Automation Articles