Zonal Recognition
SmartZone OCR includes the following functionality for OCR recognition:
Recognition through Character Sets
The ability to recognize machine printed text based on pre-defined CharacterSets. These character sets include uppercase, lowercase, and mixed case alphabetic, alphanumeric, digits, currency symbols, combined currency and digits, symbols, phone, and punctuation characters. Language may be specified, or the default value of WesternEuropean which includes all supported languages may be used. Optimal recognition will be obtained by using the smallest language set that includes all expected characters.
The ability to limit possible recognition results within defined character sets as needed is also provided. Recognition will be improved by constraining the list of possible results as closely as possible, while still including all possible results.
See the section Define and Edit Character Sets for more information.
Recognition through Field Type
The ability to specify the format of the field allows for greater accuracy. SmartZone OCR supports predefined field types of currency, date, time, email, URL, Social Security number, and United States phone number.
-
If the field is of a known format and not general text, then recognition can be improved by attempting to recognize the image to the format.
-
If the data is in a different format, then SmartZone OCR allows custom format or “masking” using regular expression.
-
If the data is a list of known entries, such as, an employee list or part numbers, then SmartZone OCR supports this with DataValidationList.
See the sections Define and Edit Field Types, Define and Edit Data Validation Lists, and Define and Edit Regular Expressions for more information.
Confidence Values and Results
Recognition results and confidence value indicators are provided in all output.
See the Confidence Values and Determine Results sections for more information on confidence and results.
Ability to Modify Results of Recognition
The ability to modify character recognition results is available to increase accuracy through the CharacterResult class.
| Use the ... | To ... |
| Text property | modify the CharacterResult. The Confidence will be set to 100 for the CharacterResult. |
| DeleteCharacter method | remove a character within the TextLineResult. |
| DeleteTextLine method | remove a text line from the TextBlockResult. |
| RejectionCharacter property | control which character is used in the output text when recognition cannot determine a value. Use the rejection character whenever the confidence for a character is below the MinimumCharacterConfidence property value. |
