AnalyzeImage

This action converts the Image file (.tif) that represents the current page to a CCO file for the page. The new CCO will not contain any recognized text. The CCO contains geometry information that can be used to identify a page using a fingerprint.

Member of namespace

SharedRecognitionTools

Syntax

bool AnalyzeImage ()

Parameters

None.

Returns

False if the ruleset with this action is not bound to a Page object of the Document Hierarchy or if the CCO cannot be created.. Otherwise, True.

Level

Page level.

Details

A ruleset with this action should be bound to a Page object that represents an application's source page. The action is not required if full-page recognition takes place using actions such as RecognizePageOCR_S or RecognizePageOCR_A.

Typically image enhancements to adjust image quality and rotation must be performed prior to calling AnalyzeImage to create the fingerprint CCO which will be used to match subsequent documents of the same page type. For example, if an image comes in rotated, it will not match the fingerprint of the same type that is in the correct orientation.

Example

RotateImage()
AnalyzeImage()
SetProblemValue(0.5)
SetSearchArea(0.5)
FindFingerprint(True)

This sequence rotates the image to get the image into the correct orientation then generates a CCO file for the current page. Finally, the sequence attempts to match the current page with a fingerprint. (For more about the matching process, see the descriptions of the <b>AutoDoc</b> actions.) Note that recognition libraries can also rotate the image and may work more reliably on some pages because RotateImage rotates based on image geometry while recognition engines rotate based on the target language.