› Capture Profile Configuration › Recognition › Patch Codes

Patch Codes

Select Enable patch code detection for the program to detection patch codes and recognize patch code types (I, II, III, IV, VI, or T) during the capture process.

To know what to do after detection, see Separation.

The following options are only available if patch code detection is enabled.

Select the Recognition engine to be OpenText or Aspose.

Clicking Image Processing opens the Image Processing Settings dialog box, where you can add image processing commands. See Image Processing.

Clicking Preview opens the Patch Code Recognition Preview dialog box, where you can select an image and view the patch code on that image.

  • Image Processing: See Image Processing.

  • Reset Image: Removes the image that was selected or loaded with the appropriate toolbar icons.

Toolbar icon Description
Save

Saves the current patch code type.

Select Image

Prompts you to select a sample image page from a file containing patch codes (to determine the types of patch codes present) and check for functionality.

Capture Image

Prompts you to capture a sample image page containing patch codes (to determine the types of patch codes present) and check for functionality.

Recognize

Lists the detected patch codes under Recognition Results in the order they appear on the viewer pane.

OpenText engine options

These options are only available if the selected recognition engine is OpenText.

You may further optimize the patch code recognition accuracy and performance by specifying the reading quality of patch codes:

  • Most Accurate: A slower process with more accurate results.

  • Fastest: A faster process with a bigger margin for error.

Depending on the page orientation, select the appropriate Patch Code orientation to enhance the speed of patch code recognition. The recognition engine attempts to recognize patch codes only in the given directions. If you are unsure of the patch code capture direction, select all directions to ensure that all patch codes are read correctly.

Aspose engine options

These options are only available if the selected recognition engine is Aspose.

You may further optimize the patch code recognition accuracy and performance by specifying the reading quality of patch codes:

  • High Performance

  • Normal

  • High Quality

  • Maximum Quality

Scan window size: The size of the scan window.

Options

Skip recognition on back pages: Skip patch code detection on back pages, leading to increased performance during capture.

Zones to Recognize

Select to detect the entire page (default) or define a zone for the recognition, which would increase performance. Patch codes outside the selected zone are not recognized. Click Define Zones to configure a zone.

To define zones, zoom in to the desired area on the viewer pane (with the Zoom In toolbar icon), then select the Draw Zone toolbar icon and draw the zone.

Drawn zones are listed in the Zones pane. Name the drawn zone and fill out which page of the document it is expected to be found on. The zones in the list can be used anywhere in the program where the user can define zones.

The page and resolution of the template are displayed at the bottom of the screen. They must match the actual page and resolution at capture time.

In the Zones pane, a set of icons are available.

Icon Description
Edit Zone

Opens the OCR Smart Zone Configuration dialog box and allows editing of child zones, anchors, and more.

Delete Zone

Deletes the zone.

If a zone is in use by any capture profile, the zone cannot be deleted.

Image Processing

Applies image processing to the zone template image.

Zone Preview

Previews OCR (printed text), ICR (handwritten text), MICR (E-13B), or MICR (CMC-7) results in the selected zone.

Ungroup Child Zones

Ungroups a cluster of child zones contained within the selected zone (child zones are used for OMR purposes).

Zone Advanced Settings (OmniPage)

Opens a dialog box, where you can configure the advanced settings of the zone, depending on the used recognition engine. This description is for the OmniPage engine.

Enable spell check: This feature accounts for everyday norms when generating results. For example, spell check ensures a number such as 34567 would not be read as 34S67, as it is unlikely that the letter S would appear within a list of numbers. If your images are of high quality (computer-generated) and a zone contains data that breaks everyday norms (in the example above, contains letters and numbers) disable this option for the zone to improve your OCR results.

Enable vertical dictionaries: Vertical dictionaries are seen in the legal, medical, and financial fields, and the OmniPage engine supports several languages with vertical orientations.

Text direction: Specify the direction of the text for the OmniPage OCR engine to read from:

  • Horizontal: Standard text on a horizontal plane, from left to right.

  • Auto: Automatically determine the orientation of the text during recognition.

  • Left: Limit OCR to only the standard horizontal plane from left to right.

  • Right: Read text from right to left on the horizon plane.

  • Vertical text (CCJK only): Support for Chinese, Cantonese, Japanese, and Korean (only) in vertical text orientation.

Auto-detect hand-printed text (ICR): Automatically detect hand-printed text (ICR) and adjust recognition settings accordingly. This feature is an enhanced version of the legacy feature "Machine Print Font Type - Unknown". Only subregions classified as machine print are further processed. Hence, there is some risk that low-quality machine print subregions are missing in the final result, for they might have erroneously been classified as handwriting.

Enable dynamic bounds for this zone...: Automatically resize zones based on the dimensions of the captured page. For example, if a zone is defined to encompass the top half of an 8.5" x 11" page, and an 8.5" x 5" page is captured, the zone's height is automatically adjusted to cover only the top half of the page.

Zone Advanced Settings (RecoStar)

Opens a dialog box, where you can configure the advanced settings of the zone, depending on the used recognition engine. This description is for the deprecated RecoStar engine.

Enable OCR logical context filtering: OCR logical context filtering enforces logical rules when generating results. For example, the number 4567 would not be read as 4S67, because it is unlikely that the letter S would appear in a list of numbers. If your images are of high quality (computer-generated) and a zone contains data that breaks logical rules, disable this option for the zone to improve your OCR results.

Enable OCR Trigram mode: Trigram post-processing uses language model statistics to try and correct recognition results in natural language sequences. Disabling Trigram mode can improve recognition results on zones that contain random sequences of letters and numbers (for example, part numbers, invoice numbers).

Machine print font type: The options include Unknown and Machine Type.

If Machine Type is selected, machine-print recognition is activated. All text detected in the input area is treated like machine print (handwriting included).

Selecting Unknown activates automatic handwriting detection. Only subregions classified as machine print are further processed. Hence, there is some risk that low-quality machine print subregions are missing in the final result, as they might have erroneously been classified as handwriting.

Machine type pitch: The options include Unknown and Variable.

In rare instances, the default value Unknown can cause misreads in zones that contain a mixture of numbers, characters, and letters. For example, "25-IN" would be read as "254N". In these cases, switch the machine type pitch to Variable.

Enable dynamic bounds...: Select this option to automatically resize zones based on the dimensions of the captured page. For example, if a zone is defined to encompass the top half of an 8.5" x 11" page, and an 8.5" x 5" page is captured, the zone's height is automatically adjusted to cover only the top half of the page.

For a full breakdown of the zone editing process, see OCR Smart Zone Configuration.