Labeled OMR

Can be one of the following values:

The OmrMode enum determines the extraction logic and output format for OMR results:

Usage Guidance

• Choose the mode that matches your form's design and data requirements.
• The selected mode affects how results are grouped, formatted, and which properties (such as SeparatorString or ValueIfChecked) are used.

Examples

• Use for radio button groups or single-choice questions.
• Use for checklists or multi-select questions.
• Use for yes/no, present/absent, or on/off checkboxes.

Can be one of the following types:

If Label Sets are in use, this property should be left empty, and checkbox labels should be defined in the Label Set.
Labels may be matched individually or as a group, depending on the value of the "Group Match" property.

This property can also be left empty if List Values have been configured on the parent Data Field. In this scenario, a temporary List Match extractor will be created and used.

If Group Match is False, the extractor should be configured to match individual labels. For example, Pattern Match looking for yes|no|undecided, or Label Match looking for the list "Yes, No, Undecided". When using this mode, the "Minimum Label Count" property should be set, to tell the extractor how many labels are supposed to be in a group.

If Group Match is True, the extractor should match an entire group of labels, and return the individual box labels as children.
This can be done by using Pattern Match to define a regular expression with named groups, or by using a Data Type with Array, Ordered Array, or Pattern-Based collation. When using this mode, the extractor is guaranteed to hit a specific number of labels, and there is no need to specify a Minimum Label Count.

Can be one of the following types:

Specifying a header is optional, and is only required for disambiguation purposes. For example, if the document contains 5 groups of checkboxes with the labels ☑ Yes ☐ No ☐ Undecided, the header extractor can be used to select the group under the heading "Attending next semester?".

If Label Sets are in use, this property should be left empty, and a header label should be defined for the field instead.

When enabled, the Label Extractor should be configured to match a group of labels as a single result, with individual box labels as children.

• Use this for forms where labels are grouped together and can be matched as a block.
• When disabled, the extractor should match individual labels, and the "Minimum Label Count" property should be set.

This property only applies when using a local Label Extractor (rather than Label Sets), and the Group Match property is set to false.

• Set this to the expected number of labels in a group to ensure only complete groups are matched.

The "bounding region" is defined as smallest rectangle which can enclose the labels. Stated another way, it is the geometric union of all label rectangles (locations).

Noise characters are letters or digits which are present in the bounding region, but not part of the key or the value. Non-word characters such as punctuation, whitespace, and control characters are not counted as noise characters.

Setting this property to zero will invalidate results with any noise characters. Results will not be checked for noise characters if this property is not set.

Enable this feature when multiple labels with the same text appear on form pages, and cannot be organized into groups accurately by location alone.

• This helps prevent incorrect grouping by using intersecting lines as additional grouping cues.