In the imaging industry, the term "Barcode Recognition Rate" refers to the likelihood that a symbol is (a) found on the image and (b) is correctly decoded. Thus it is a measure of accuracy rather than of speed. ClearImage barcode recognition achieves 100% recognition on the vast majority of images.   A number of factors can adversely affect the ability to find barcodes on the image or to decode them correctly.   The recognition process consists of steps, described below, showing the impact of impairments at each step:

1. Encoding - Barcode image generation software, also called Barcode Fonts or Barcode Printing software, translates input data into barcode image.   It should perform the encoding correctly according to all aspects of standard. There are issues with the spacing and with the error correction computations in some cases.
2. Printing - The symbol image is output to paper.   The printing resolution should be sufficient to make sure that printed bars and spaces width are correct multiples of module size.   Printing at insufficient resolution introduces printing digitizing error. Note that the printed symbol must be scanned (at 100 or 200 dpi), and then recognized. The module size must be evaluated. Typically, the module size should be 30-40 mil (~0.1 mm). A 600 dpi printer should, therefore, print 18-24 print dots per module. If the symbol is too large, there are other techniques for reducing its size. Contact Inlite Support for advice.
3. Handling - Document handling by people or machines might pollute the barcode image with pen marks, scrapes, stray marks, stamps and other artifacts or dirt.
4. Scanning - Scanning transfers barcode back into electronic form, and it introduces scanning digitizing error. The combination of printing and scanning digitization errors, especially when their dpi values are closely related, can render the apparent barcode image unrecognizable.
• Scanning done at too low to high brightness might lead to broken or touching bars.  In extreme cases, the bars might disappear or merge completely, destroying the symbol.
• Paper "chewed" by scanner or fax machine introduces non-linear distortion to page image.   If such distortion intersects barcode portion of the image, it affects the ability to recognize the symbol.
• Fax transmission at standard resolution is uneven and rough, producing an image that is 96 dpi by 192 dpi. ClearImage Repair must be used to correct his assymetry 
5. Recognition - The barcode recognition process, performed by ClearImage engines, recovers the input data from the barcode image.  Measuring module size and barcode geometry it interprets the results according to the relevant standard to obtain input data. In the real world, barcode images are not perfect and ClearImage engines will apply various algorithms to compensate for errors and distortion. 

To improve recognition rate read:

• Barcode printing optimization
• Barcode scanning optimization

© Inlite Research, Inc. 1992-2005 All Rights Reserved.