Add an OmniPage page recognition profile
Use the procedure in this topic to add an OmniPage page recognition profile.
This recognition profile has several unique settings related to the OmniPage recognition engine.
-
On the
Project tab, in the
Configuration group, select
Project Settings.
- Select the Recognition tab to view the recognition settings.
-
Select the
Page Profile button.
A properties window is displayed.
-
From the
Page Recognition Method, select
OmniPage
2025.1.
The properties displayed on the window automatically update to the selected recognition method.
-
In the
Languages list, select one or more languages that your documents will use.
This list enables you to select one or more languages from a comprehensive list of available languages. You can combine any languages. The English language only is selected by default.
As you select additional languages, the language count in the header increases. Not all languages are visible and this ensures that you are aware that some languages lower down in the list are selected.
If you have purchased the Asian Plus OCR add-on, additional languages are supported.
If you select English plus one additional language, that other language is given priority during recognition. For example, a selection of English and Greek assumes the majority of the documents are in Greek with only a few words in English.
If there are many documents, some mostly English, and others mostly Greek, then it is recommended to pre-sort the documents and then use different projects for production. Alternatively, use a single project and classify the documents into the correct language before recognition is executed, and then use a re-read profile for the dominant language.
-
In the
General Settings group, configure the following settings:
Setting
Description
Word separation characters
Use this field to define what characters may separate words. (Default: /:()-#)
For Chinese and Japanese characters, the separation characters defined above, or a space must be present at the end of a character, in order for it to be recognized as a word. Sequential characters with no spaces or separation characters are not recognized as individual words.
Correct split numbers
Combines number values that are physically close together and are incorrectly split during recognition. (Default: Cleared)
For example, an amount of 32,49 is incorrectly separated into two pieces of data. This setting enables you to correct this split.
Recognition mode
Select the Recognition mode from the following values.
- Fast:
-
Choose this value to prioritize speed of recognition over the recognition accuracy. If you choose this value, the recognition process takes less time, but the overall recognition accuracy may decrease.
- Balanced:
-
Choose this value to balance the priority of the recognition accuracy with the speed of recognition. (Default: Selected)
- Accurate:
-
Choose this value if you want to prioritize the recognition accuracy over the speed of recognition. If you choose this value, the recognition process takes longer.
MRZ Mode
When selected, a specialized recognition engine processes the page for MRZ text and the Print Type is set to OCR-B automatically. This is because OCR-B is the font used by MRZ.
Any non-MRZ text on the page is likely to have low recognition accuracy. Because of this, if you want to recognize other non-MRZ text on the document, consider using zone recognition.
Image transfer mode
Choose how images are transferred for this recognition method. You can choose between the following values.
- Memory:
-
Choose this value to use the binary image that is already stored in memory. (Default: Selected)
- File:
-
Choose this value to provide the file path so it can be accessed directly by the recognition engine. If chosen, any image pre-processing performed on the image is ignored during recognition.
When the File value is chosen for the Image transfer mode, you cannot use the OmniPage page profile for the Mixed Print page profile. If an OmniPage page profile is already used by a Mixed Print page profile, then selecting the File transfer mode removes that OmniPage page profile from the corresponding Mixed Print profile, and an error message is displayed.
A Mixed Print page profile is not able to process images when their image transfer mode is file-based. This is because in mixed recognition, two images are created in memory, one for each recognition print type configured in the Mixed Print page profile.
-
In the
Dictionaries group, configure the following settings:
Setting
Description
Language dictionary
Runs spell checking during recognition for all of the following languages.
- Brazilian
- Catalan
- Czech
- Danish
- Dutch
- English
- Esperanto
- Finnish
- French
- German
- Greek
- Hungarian
- Italian
- Norwegian
- Polish
- Portuguese
- Russian
- Slovenian
- Spanish
- Swedish
- Turkish
User dictionary
Specify your own dictionary that is used to aid spell checking during recognition. It is important to know that a misspelled word in the recognition results is not always replaced with a corresponding entry from a user dictionary. (Default: None)
Format the dictionary so there is a single word per line in the dictionary file.
Once selected, browse to select a dictionary.
User dictionaries are not supported for the Asian Plus OCR add-on languages.
Business terms dictionary
Uses a dictionary and then select a business terms dictionary from the list. This dictionary can improve spell checking during recognition using the contents of the selected dictionary. The following dictionaries containing business terms are available. If there is not a suitable setting here, use the User dictionary setting above and specify your own list of terms. (Default: None)
-
English Financial Dictionary
-
Dutch Medical Dictionary
-
English Medical Dictionary
-
French Medical Dictionary
-
German Medical Dictionary
-
Dutch Legal Dictionary
-
English Legal Dictionary
-
French Legal Dictionary
-
German Legal Dictionary
-
In the
OmniPage Raw Results group, configure the following settings:
Setting
Description
Provide page elements
Provides raw page results by specifying a format and scope. (Default: Cleared)
Scope
Available only when the Provide page elements setting is selected.
Select a value for the scope that is included in the raw results.
- All elements:
- When selected, all information from all pages of a document is provided. (Default: Selected)
- Tables only:
- When this value is selected, only table information is provided.
Data format
Available only when the Provide page elements setting is selected.
Select the format for how the raw data is recorded. Select from:
-
XML (Default: Selected)
-
JSON
- Optional. Rename the new profile so it has a descriptive name.
- Optional. Select OK to close the Project Settings window.
- Save the changes to your project.