Text field recognition options

Recognition properties can be specified on the Recognition tab of the field properties dialog box (click the Properties... item on the field's shortcut menu to open the properties dialog box). Correctly determined field values will improve recognition quality and reduce the likelihood of errors.

In the Filling type field, select the desired recognition mode: Standard recognition, Do not recognize or Script recognition.

For the Standard recognition mode, specify the following:

  • ICR (hand-printed) - Enable this option if the field contains hand-printed static text. As the style of writing can vary significantly from country to country, you need to specify the appropriate country.

    Show sample writing styles for digits

  • OCR (printed) - Enable this option if the field contains printed static text. Select the print style from the drop-down list (typographic, dot-matrix printer, typewriter, etc.). See also Supported text types.
  • Advanced - Use this option to select several text types or to specify a custom text type. This option also allows you to load a pattern file in PTN format. You can create user patterns in ABBYY FineReader Professional/Corporate Edition 9.0 (or earlier) or in ABBYY FineReader Engine 9.0 (or earlier).


Pattern

To load a pattern, click Modify.... In the dialog box that opens, select Use pattern and specify the path to the file.

  • Marking type - Select the marking type; to simplify your choice, select samples from the drop-down list. If the marking disappears during scanning, your marking type is monospaced (gray boxes). If the marking does not disappear during scanning and is divided into cells for character spaces, you must enter the total number of cells. When the program detects fields with such marking automatically, the number of cells is also determined automatically.
  • Letter case - Select the case of letters in the field. If both lowercase and uppercase letters are possible, leave the Auto option enabled.
  • Orientation - Specify the text orientation.
  • Direction of CJK text - Select the reading direction to be used for OCR of texts in Chinese, Japanese, or Korean. Possible options are Auto, Horizontal Script, or Vertical Script. Auto is selected by default and is the recommended option for fields that do not contain any CJK text.
  • Enable the One line option for one-line fields.
  • For fields whose value will always consist of a single word, enable the One word option. You can also enable this option if you want to apply a regular expression to the entire field irrespective of the number of words in it.
    Note: With the One word option enabled, it is not recommended to have expressions in the custom dictionary that contain the space character.

Specify image pre-processing settings:

  • Invert inverts image colors and brightness during recognition (this inversion is temporary and only affects recognition; original image colors will be retained in the output file).
    • Autodetect automatically detects the text color and background color and inverts them if necessary. This is the recommended setting for documents that contain both light text on a dark background and dark text on a light background.
    • Invert inverts images completely.
    • Don't invert keeps original colors (this option is enabled by default).
  • Remove texture removes texture.
  • Despeckle - Enable this option to remove garbage from the image.
  • Clear the garbage of specified size only - Enable this option if you want to remove garbage of only specified size. Specify garbage size. If this option is disabled and the Despeckle option is enabled, garbage size will be selected automatically.

Select the Do not recognize mode if the field cannot be recognized for some reason (for example, if the text in the field consists of fused letters), and the Operator is to enter its value manually. In this case, you do not have to configure other recognition properties because this field will not be recognized, and the Operator will be asked to enter the field value during verification.

Select the Script recognition mode if you want to use a custom recognition or initialization algorithm. For more details, see Custom recognition script.

25.09.2020 9:24:45


Please leave your feedback about this article