The OmniPage Capture SDK architecture is designed to accommodate multiple image processing technologies through four main subsystems:

• An image input subsystem for scanning or importing images.
• An image preprocessing subsystem for improving image quality prior to recognition.
• A recognition subsystem that provides multiple recognition technologies for image processing.
• An export subsystem to format the output from multiple recognition modules into a common format for conversion to popular word processing formats or ASCII.

Interfaces

Two programming interfaces are available with the OmniPage Capture SDK:

• C API
  The C API allows control over image input, image preprocessing, recognition, and output and supports image processing on a page basis.
  
  
• ActiveX
  An ActiveX interface is provided for Visual C++ and Visual Basic programmers. This interface includes all of the functionality of the C interface and offers document processing capabilities allowing programmers to create solutions that manage documents more efficiently. This interface also expands the support of modern development environments, including managed environments like VB.NET or C#.
  
  
• Professional Toolbox
  In conjunction with the ActiveX interface, a set of controls, collectively called the Professional Toolbox, is available as an add-on module. These pre-made controls allow developers to reduce development time and speed time-to-market by supporting easy creation of controls for:
  
  
• Image viewing
• Zone content validation
• Image thumbnail viewing
• Text verification and editing
• Display statistical information and a draft of the document
• Provide details and progress about the workflow being executed on the system
• Create OmniPage compatible workflows
• Access and change output converter settings
• Display and edit form fields and attributes

Image Input

The image input subsystem provides scanner and image conversion interfaces. In addition to TWAIN scanner support, Nuance supplies its own set of driver routines for Kofax, HP, Bell & Howell, Ricoh, Fujitsu and Panasonic scanners.

Both color and grayscale images can be handled by the OmniPage Capture SDK and application developers can send images from memory to the preprocessing and recognition processes.

Input conversion for TIFF, TIFF/JPEG, TIFF-FX, PCX, DCX, BMP, ADF, JPEG, PNG, PaperPort MAX and PDF image formats is available.

Note: For ISIS support, scanning requires an ISIS driver to be installed on the computer.

Image Pre-processing

Image correction and pre-processing can greatly enhance the quality of the image to achieve more accurate recognition results. Pre-processing capabilities offered in the OmniPage Capture SDK include:

• Rotate (90, 180, 270 degrees)
• Deskew (auto and programmed)
• Invert (auto and programmed)
• Despeckle
• Resolution Enhancement

An interface for integrating additional image preprocessing technologies is also available and extends the system's functionality by permitting customization of the system's image processing capabilities.

Recognition Module Management

The OmniPage Capture SDK's component manager supports the integration of 12 individual recognition modules into the Developer's application. Modules for machine print OCR, ICR (handprint OCR), Barcode, OMR (Checkbox), OCR-A, OCR-B and E-13B (MICR) are provided.

An interface is also provided for developers who want to incorporate additional recognition technologies into their application. This interface provides the mechanism to pass images, receive recognition output and pass configuration commands to the desired recognition module.

Asian OCR is supported in the OmniPage Capture SDK. It can recognize Simplified and Traditional Chinese, Japanese, and Korean with full layout retention.

See Asian OCR Support for more information.

Output Processing

The OmniPage Capture SDK's output processing subsystem is responsible for taking output from the recognition modules and converting it into a desired format.

A wide range of image and application formats are supported including BMP, GIF, TIFF, PDF, HTML, Microsoft Office formats, XML, Open eBook and more.

PDF output is supported in four formats including:

• PDF Normal (text only)
• Normal with image substitutes
• Image only
• Image on text

See Integrated PDF Toolkit for more information.

The OmniPage Capture SDK is available in 3 configurations with 2 optional add-ons:

The Professional Recognition Kit

• C Libraries and an ActiveX interface
• 2 Pre-Made voting OCR (machine print) recognition modules
• Access to 3 Individual OCR engines for application optimization
• OCR-A, OCR-B, E-13B (MICR)
• 2 ICR (handprint) recognition modules
• OMR (Checkbox)
• Barcode recognition

The Professional OCR Kit

• C Libraries and ActiveX interfaces
• 2 Pre-Made voting OCR (machine print) recognition modules
• Access to 3 Individual OCR engines for application optimization
• OCR-A, OCR-B, E-12B (MICR)

Asian OCR Kit

This kit provides support for Japanese, Traditional and Simplified Chinese, Japanese and Korean OCR with full layout retention and searchable PDF output.

Add-On Options

• PDF Output Module
  This optional add-on provides PDF export filters for output in PDF Normal, Normal With Image Substitutes, Image Only and Image On Text formats.
  
  
• Professional Toolbox
  This optional set of OCX controls provides pre-made scanning, image clean-up and GUI elements for Microsoft Visual development tools allowing developers to easily add image viewing, zone content validation, thumbnail viewing, text editing ad text verification functionality to applications.