Working with OCR Pages

After starting the OCR engine, you can begin working with the document page(s). Pages can work with or without An OCR document.

The LEADTOOLS OCR methods provide support for the following when working with OCR pages:

• Creating one or more pages from a RasterImage directly without an OCR document
• Creating one or more OCR documents
• Adding or removing pages to an OCR document
• Getting information about one or more pages
• Exporting pages
• Updating pages

As described in Programming with Leadtools .NET OCR, an IOcrPage page can be created directly using CreatePage without using an IOcrDocument object. These pages can be zoned and recognized and the OCR results can be obtained directly using GetText or GetRecognizedCharacters.

If the page are to be saved to a final document such as PDF or DOCX, then an IOcrDocument instance is required.

An instance of IOcrDocument contains the pages of a document. You can create a new OCR document using the IOcrDocumentManager.CreateDocument method. This method allows creation of memory-based or file-based documents.

Each OCR document can have one or more pages (IOcrPage objects). Each IOcrDocument contains an IOcrDocument.Pages property of type IOcrPageCollection that you can use to access the pages of a document.

IOcrDocument through the IOcrDocument.Pages property holds a collection of IOcrPage objects. Each of these IOcrPage objects contains the raster image used to create it (the image used when the page is loaded or added) and a group of OCR zones for the page either added manually or through auto-zoning.

The IOcrPageCollection interface implements standard .NET Collection{T}, IList{T}, and IEnumerable{T} interfaces and hence, you can use the member of these interfaces to add, remove, get, set and iterate through the different pages of the document if the document was memory-based.

For file-based document, adding a page to the document involves taking a snap shot of the current recognition data and store it internally. The page itself is not added to the collection and is not required to stay in memory. The collection is a read-only view on the document and the user can only add new pages and not remove or iterate through them

The following list contains the major functionality of the IOcrPageCollection interface:

• Add new pages to the document from raster image files. These files can be in disk files, a .NET stream (memory or otherwise) or even in a remote URL. The following table lists all the page addition method groups:

• Export pages from the OCR document to raster image files. You can save the pages in disk files, .NET streams or as a single or multi-page RasterImage object with any of the file formats supported by LEADTOOLS. Exporting a page is supported by memory-based documents only. The following table list all the page exporting method groups:

• Perform auto image preprocessing on a single or multiple pages in the document through IOcrPageCollection.AutoPreprocess. These methods provide a shortcut for iterating through the pages in the collection and calling IOcrPage.AutoPreprocess on each page. This is supported by memory-mode document only.
• Perform auto-zoning on a single or multiple pages in the document through IOcrPageCollection.AutoZone. These methods provide a shortcut for iterating through the pages in the collection and calling IOcrPage.AutoZone on each page. This is supported by memory-mode document only.
• Recognize a single or multiple pages in the document through IOcrPageCollection.Recognize. These methods provide a shortcut for iterating through the pages in the collection and calling IOcrPage.Recognize on each page. This is supported by memory-mode document only.
• Perform high level "fire and forget" approach to OCR recognition with multiple-threads support through IOcrAutoRecognizeManager
• Recognize single or multiple files and save recognition results to a file with a single function by calling IOcrAutoRecognizeManager.Run, IOcrAutoRecognizeManager.RunJob or IOcrAutoRecognizeManager.RunJobAsync methods.