SignatureValidation actions
Use the Signature Validation actions to detect and validate signatures on a document or zone.
Object Datacap.Libraries.ParascriptFx.SignatureValidation
Default instance name: Datacap.Libraries.ParascriptFx.SignatureValidation Library to perform signature validation.
- Validate signatures on a page or field zone
- Detect signatures on a page and validate them. * This feature is provided as a "Preview" feature
- The page must be a US Check. The signature can be validated with the ValidateSignature action.
- Detect and validate signatures on a page. The action SearchAndValdiateSignature attempts to identify signatures on a page and validate the areas that it identifies as a signature.
The signature validation action library provides the ability to validate a zoned signature or to validate a signature detected on the page. It is recommended to use field/zoned signatures as often as possible. A zoned signature provides the best results for signature validation. If it is impossible to zone a signature, signatures can be automatically detected and validated at the page level. Automatic detection can sometimes misidentify or miss signatures on a page. See the action SearchAndValidateSignatures for details.
Signature validation requires a minimum number of signature reference images for each signer that is validated. The help for actions CreateNew and SetSignatureReferenceFolderPath covers the minimum number of reference images required for each signature and how to use them to validate signatures.
Image Requirements
Images need a minimum dpi of 200 and a maximum dpi of 300. Images outside this range are not supported. The image must be a TIF that is black and white or an 8-bit grayscale image. Automatic page level signature detection works best clear area around the signature.
Images need to be isotropic, meaning that the X dpi (dots per inch) equals the Y dpi. If some ingested images may be non-isotropic. For example, fax images usually have a different X and Y dpi. Use the "EqualizeUnbalancedImage" action to convert the image so the X and Y dpi are the same prior to fingerprinting, recognition, etc.
As with text recognition, signature detection and validation is not 100% accurate. Applications need to handle incorrect detection or validation. Clean images help improve results. Use a lossless compression for your images, such as LZW or Fax compression. Do not use a lossy compression, such as JEPG, for any pages that are used for recognition. It is recommended to review the "Best Practices for optimal text recognition in IBM Datacap".Action Overview
The following is a brief overview of how the signature validation actions can be used within an application.