Nuance - Document Imaging Solutions

What's New in Version 16

Significant accuracy and speed improvements plus robust new features and continued innovation make version 16 the most comprehensive OmniPage Capture SDK release ever. Refreshed and updated with significant improvements without changing the API from version 15, it’s easy to upgrade your application’s capabilities with the OmniPage Capture SDK 16. Features aside, the performance improvements alone insure that the OmniPage Capture SDK is destined to remain the #1 OCR toolkit in the world and the continued choice of imaging technology providers among nearly all commercial imaging vendors

Increased Accuracy
Already judged by industry experts to be the most accurate OCR solution available, the new OmniPage Capture SDK release delivers an increase in accuracy over already impressive previous versions due to pre-processing binarization that improves image quality before the OCR process occurs. Best of all the new release delivers superior accuracy with improved throughput where alternative SDK products exchange accuracy for slower throughput – making the OmniPage Capture SDK the ideal solution for high-volume document capture applications.

The Fastest OmniPage Capture SDK Ever!
Now work up to 67% faster than ever before. Improved hyper-threading accounts for 20% of the all new speed increase in the OmniPage Capture SDK. New parallel processing algorithms developed in cooperation with Intel segment multi-page document conversion tasks over multi-core processors for unrivaled performance. Where lesser OCR SDKs will sacrifice accuracy for speed, OmniPage leads the industry on both counts.

Unrivaled Layout Retention

• Superior Table Recognition — Table accuracy is the most requested improvement in the conversion of paper or PDF files to fully editable documents. Table analysis has been improved to better cope with the large variety or ways a table can be created and formatted with text, numbers and borders. OmniPage Capture SDK delivers accurate results for table cells and rows containing multi-line texts within a cell, correcting splitting methods for incompletely filled tables, improving the decision-making on separator lines and preventing tables of contents (TOC) from being recognized as tables.
• Support for Vertical Text — Vertical text is most often used as labels for graphics, inside tables, or for naming document pages. Automatic detection and zoning of left or right rotated vertical text including detection inside table cells is now available in the OmniPage Capture SDK. The Professional Visual Toolbox has also been upgraded to accommodate vertical text so that plug-in zoning tools and text-editor can be easily added to your application.
• Fast and Accurate Font Matching and Spacing — Using an incorrect font or using word and character spacing incorrectly can affect the layout of a document significantly as text lines overflow or become artificially compressed in a column. Both font attribute recognition and font mapping are improved for Western languages in the OmniPage Capture SDK while the font matching is better aligned with language choice in the Asian kit for accurate layouts.
• Automatic Margin Correction and Page Consolidation — Uniform page margins can become distorted, clipped or exaggerated through scanning causing uneven margins from page to page in the final output. The OmniPage Capture SDK 16 examines incoming pages, can determine if that they have a similar text area and layout, and automatically produces the optimum page margins that most closely match the original for final output.
• Large Title Recognition — Large titles pose recognition challenges because they can be overly stylized resulting in distorted characters, blend with multiple background colors or run into pictures. Innovative new algorithms identify page areas that could be either titles or pictures and examine them more carefully, searching for letter-like shapes and getting the recognition correct.
• Special Recognition of Legal Document Elements — The OmniPage Capture SDK 16 introduces unique algorithms to better handle specialty legal documents often called pleading documents. Legal recognition addresses pleading numbers (line numbers at the edge or margin of the document), document identifiers (naming the plaintiff, defendant and case number), and the optional removal of page headers, notes, stamps and markups. Combined with other high value features in the OmniPage Capture SDK you can now build specialty applications for profitable legal and government markets.

Superior Parsing of Text-based PDF Documents
Text-based PDF files already contain information necessary to reproduce a PDF into an editable format so no OCR process is necessary. Alternative OCR SDK products that cannot accurately read all this information will treat the PDF as an image and apply a more inaccurate OCR process. The OmniPage Capture SDK is unique in being able to accurately read a text based PDF and converting it into editable formats that look just like the original.

Integrated Asian OCR Kit
Recognition and associated services are provided for Traditional Chinese, Simplified Chinese, Japanese and Korean using a common API with the Western kit. This means Asian OCR is available through a vastly improved interface compared to the separate Asian version 15; in place of its 40-50 functions, there are now over 200 accessible functions for control and operation consistent with the Western Kit.

Enhanced Support for 24/7 Applications
The OmniPage Capture SDK 16 introduces new features to make high-volume and unattended 24/7 processing more robust. A new self-standing sample application and Windows service demonstrate how an application can be properly architected for 24/7 use. It is now possible to separate the OmniPage Capture SDK code from the user application process delivering unrivaled flexibility and robustness for unattended high-volume tasks. Updated coding guidelines and samples ensure reliable operation.

True Office 2007 Output Including DOCX, XLSX, PPTX and XPS
Nuance and Microsoft have had a close relationship on OCR and document conversion for many years, including a collaboration that resulted in PDF Converter, the first-ever solution that accurately turned PDF into Microsoft Office Word documents. The collaboration has continued through the development of export filters for OmniPage that leverage the Ecma Open XML file formats used in 2007 Microsoft Office system applications, including Office Word (.docx), Office PowerPoint (.pptx) and Office Excel (.xlsx). Nuance has also added robust support for the XML Paper Specification, XPS, which enables document sharing and archiving with Windows Vista. Now you can deliver these innovations to your customers through the OmniPage Capture SDK.

Amazing 3DC Digital Camera Image Correction Technology
The ability to perform OCR on a digital camera image was the #1 requested new feature for OmniPage 16. Now this technology is available for developers to build into their own applications with the OmniPage Capture SDK. Innovative new 3DC technology automatically adjusts for skew, waves and 3-D perspective. The results are up to 74% more accurate! So go ahead and let your customers grab a camera – with the OmniPage Capture SDK 16 it’s both easy and accurate to click, capture and convert text without a scanner.

Powerful Automatic Data Collection from Machine Printed or Text-based PDF Forms
For organizations needing a better way to implement electronic and paper forms processes, scanning and OCR alone are not sufficient. While sending PDF forms is easy, the challenge is in processing completed forms into data – a challenge that is met with the new release of OmniPage Capture SDK 16. Now developers can create applications that can batch process either machine print paper or PDF forms and let the OmniPage Capture SDK automate the data extraction into CSV files suitable for Microsoft Office Excel and database applications.

Unique ‘Scan-and-Redact’
Handling sensitive information is a serious concern for government and corporate organizations, as well as individuals concerned with names and email addresses, social security and phone numbers and other information that can be used by identity thieves. OmniPage Capture SDK 16 is the only OCR development kit to address this concern, making redaction – the removal of text from an image and PDF file – a core feature. Users can set sensitive words or phrases as an attribute prior to scanning, and any application developed with the OmniPage Capture SDK can redact on the fly. Developers can also plug-in interfaces so their users can select highlight-for-redaction, where candidate words are highlighted for review by the individual or others prior to the final redaction step. Nuance has submitted a patent application for the redaction capabilities delivered in the OmniPage Capture SDK.

Productive ‘Scan-and-Highlight’
Being able to visually scan a document and instantly identify words or phrases with a highlighted color is important for so many business, government and educational documents. OmniPage Capture SDK 16 is the only OCR development kit to address this need, making the automatic highlighting of text an easy to implement feature in your application. Your users can set words or phrases as an attribute prior to scanning, and any application developed with the OmniPage Capture SDK can highlight those words on the fly with the results appearing in PDF or Word files.

One-Button Workflow
In order to distill complex, multi-step document recognition tasks into a single click, the OmniPage Capture SDK 16 provide an inventive new Workflow interface that can be dropped into an application making the creation and use of document workflows easier than ever. In addition to all the new features and formats that Workflows now have through the SDK we’ve also added native support for the ODMA content management standard and Microsoft Office SharePoint Server 2007 in order to make storing documents even easier.

Key Features for the OmniPage Capture SDK for Windows

The OmniPage Capture SDK has a robust feature set designed to support all your imaging needs. You get the power and accuracy of OmniPage, the most popular OCR program in the world, integrated into your applications, along with top-of-the line recognition engines and extensive PDF capabilities all in one single package. The strength of OmniPage Capture SDK extends beyond our unrivaled accuracy, with additional key features designed to streamline application development and provide added value to your product.

The most accurate and robust OCR available
The OmniPage Capture SDK provides a scalable voting interface and significant throughput management capabilities. Combined with highly accurate machine print OCR (OCR, OCR-A, OCR-B and MICR), Handprint (ICR), Check Mark (OMR) and Barcode (1D and 2D) recognition engines, the OmniPage Capture SDK delivers unmatched flexibility and the most accurate results possible.

Asian OCR support
The OmniPage Capture SDK Asian OCR module supports Simplified and Traditional Chinese, Japanese and Korean. It can be used either as a standalone module or together with the Western language kit.

Support for the .NET managed environment
OmniPage Capture SDK 16 offers support for .NET, one of the leading development environments among programmers. C# and VB.NET sample recognition programs and sample viewers are included.

Multi-core and multi-thread processor support
Better multithreading and parallel processing on multi-page documents applied in the OmniPage Capture SDK lets you exploit the full potential of your processing environment. In multi-page mode, OmniPage Capture SDK 16 ran up to 43% faster than the previous version of the Capture SDK on a quad-core machine.

Pre-made user interfaces
The OmniPage Capture SDK’s Professional Visual Toolbox module provides pre-made user interfaces for creating and executing workflows, controlling scanning devices, and handling document processing capabilities. It includes visual controls for advanced image enhancement tools. Also use this module to create OmniPage compatible workflows and monitor its execution, and display and edit form attributes.

Workflow development and execution
Developers can use this feature to easily create complex image processing and OCR tasks and manages the parameters and settings. After that adding OCR to application can be juts one workflow execution call. Workflow also helps balance the load on dual core and hyper thread systems to boost performance.

Logical Form Recognition Technology
Our advanced Logical Form Recognition (LFR) automates form template creation and streamlines form processing, providing significant savings in application development time.

Throughput management
Updated throughput capabilities provide significant advantages over other SDKs, allowing developers to achieve optimal image throughput for their applications.

Integrated PDF Toolkit
Extensive PDF capabilities, including unique PDF Overlay Matching, that achieves near 100% accuracy in converting PDF documents, allows software developers to significantly reduce the cost of development and speed time-to-market. The OmniPage Capture SDK also supports output to PDF/Archive (PDF/A) format and generates Multi-Raster-Content PDFs optimized for file size and quality.

Format support
The OmniPage Capture SDK provides support for a wide range of image and application format support, including BMP, GIF, TIF, PDF, HTML, Microsoft Office formats, XML, Open eBook and more. New throughput capabilities provide significant advantages over other SDKs, allowing developers to achieve optimal image throughput for their applications. There is also support for a wide range of input/output filters, including PDF, Microsoft® Office 2007, XML and Open eBook standards.

Text-to-Speech (TTS)
The OmniPage Capture SDK is also the only OCR SDK that includes Text-To-Speech technologies. It enables applications to turn paper and digital documents into human-sounding audio files. Not only is this an important way to provide document support for disabled users, it allows everyone to save documents to files that can be played on personal computers and mobile devices, including Apple iPod.

These advanced features, along with breakthrough PDF capabilities that achieves 100% word accuracy in converting text based PDF documents, allows software developers to significantly reduce the cost of development and speed time-to-market, making the OmniPage Capture SDK the most powerful and complete imaging SDK in the world.

• Free Evaluation Version
  Register online and receive a free evaluation version of the
  OmniPage Capture SDK 16.

System requirements

The OmniPage Capture SDK recognition technologies can be accessed through a C/C++ API or ActiveX interface. Support for Windows XP and Vista application development gives developers the ability to easily create applications with a wide variety of recognition technologies using a single set of developer tools.

• Developer System Requirements
• Windows XP (SP2 or above), and Windows Vista 32-/64-bit
• Intel and AMD 32-bit and 64-bit CPUs
• Microsoft Visual C/C++ version 6.0 SP5 or above
• Microsoft Visual Studio version .NET 2003/7.1, 2005/8.0, 2008/9.0
• Microsoft .NET Framework 2 or above
• Runtime System Requirements
• Windows XP, Windows Vista 32-/64-bit, Server 2003 32-/64-bit and 2008 32-/64bit
• Intel and AMD 32-/64-bit CPUs

Product Architecture

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 text.

• Interfaces
  Two programming interfaces are available with the OmniPage Capture SDK:
• C/C++ API
  The C/C++ API allows control over image input, image preprocessing, recognition, and output and supports image processing on a page basis.
• Professional Visual Toolbox
  In conjunction with the ActiveX interface, a set of controls, collectively called the Professional Visual Toolbox, is available as an add-on module. Pre-made controls allow developers to reduce development time and speed time-to-market by allowing plug-in interfaces for your application.
  • ActiveX
    An ActiveX interface is provided for Visual C++ 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#.
  • Pre-made Controls
    - 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 TWAIN scanner and image conversion interfaces. 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 are available.

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

  See Integrated PDF Toolkit for more information.

Product Configurations

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

• The Professional Recognition Kit
  • C/C++ Libraries
  • 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/C++ Libraries
  • 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.