Open Text Capture Recognition Engine

Release Information

Quicklinks

Product Advantages Features Applications Ordering Options Technical Data

Open Text Capture Recognition Engine Version 5.0 - Highlights

New multi-language concept (*)
Supports 53 languages directly as well as many others (**)
Language dictionaries and other linguistic support for about 40 languages
Language Identification
Improved recognition
PDF output improvements
Courtesy Amount Reading
New types of 1D Barcode
MS Windows Server 2008 R2 and Windows 7 Support

For further information please contact us

(*) for the reading of machine printed documents
(**) recognition may be achieved by restricting to the necessary character sets

Open Text Capture Recognition Engine Version 4.4

Quality improvements
Open Text Capture Recognition Engine 4.4 set another highlight with an impressive increase in recognition quality especially for machine printed business documents and many functional improvements. The highlights in details are:

Significantly improved read

New machine print classifiers for several countries
Improved character separation
Improved word separation algorithms
Improved voting and postprocessing

Checkbox array processing
- Enhanced classifier includes checkmark character
- Voting algorithm for combining classification, pixel count and imaging based processing
MICR line processing
- Improved processing of disturbed lines

Recognition and Imaging Engine functionality

Regular expressions for recognition result improvement
Formatters that compares the scan results from several areas with regular expressions and prepares them in a formatted report.
Document-type recognition now using form key reading, size, layout and other features
Text document orientation detection and automatic rotation.This support applications importing TIFF or PDF Files into document archives.
Improved function for removal of dashed lines

Design Studio improvements
Significant usability improvement due to many context sensitive assistents based on experience in application development.

Full support for large forms recognition application development
Color filter editor integrated
Multi editing of parameters common to multiple fields
Table visualization of batch results
Toolbox for more efficient field definition
Improved setup image dialogue
Integrated Migration Tool for an easy migration from RecoStar 3.0 and IFG/IPPP applications and dictionaries

Compatibility and environment

All functionality is available in managed (.NET 1.1/2.0, C++, C#: RSO Interface managed) and unmanaged environments (C/C++: LGK interface)
Full capability to run and migrate existing applications (3.0 Forms generated projects) in the managed RSO Interface
Fully capability to run Design Studio Projects in the unmanaged LGK-Interface.
Supported Operating Systems
- Windows 2000 Professional/Server with SP4
- Windows 2003 Server x86 or WOW64 (32 bit emulation mode)
- Windows XP Professional (sP4) x86 or WOW64 (32 bit emulation)
- Windows Vista Business x86 or WOW64 (32 bit emulation)
- Microsoft Terminal Server

Open Text Capture Recognition Engine Version 4.0

Version 4.0 of Open Text Capture Recognition Engine provides state-of-the-art interfaces and design tools, all of which make integrating the leading recognition technology into applications as straightforward as possible. A number of migration procedures ensure that existing applications can continue to be used. The new version also extends the scope of functions by adding a new algorithm for form recognition and enhancing check box recognition. And yet another increase in recognition performance is, of course, only what has come to be expected.

.NET interface The Microsoft .Net framework increases the productivity of system integrators. Open Text Capture Recognition Engine 4.0 has implemented all relevant interfaces on the basis of .Net framework 1.1: parameter definition, runtime control and result evaluation. The results of character recognition can now also be saved in XML format.

New C++/C#/.NET interface for parameter definition, runtime and result with XML- based saving of parameter data
Saving of results data as XML data possible
Applications generated with Open Text Capture Recognition Engine 3.0 can be run; the new results data interface can process older results
Processing of image batches and multi-page documents in PDF or Multi-TIFF files is possible
Straightforward error management
Unmanaged applications can use Open Text Capture Recognition Engine V 4.0 in mixed mode, enabling a smooth transition of projects to .NET
‘Intellisense’ online help is integrated in Visual Studio
Example code facilitates application development

RecoStar Design Studio
The RecoStar Design Studio is a state-of-the-art development environment for creating
RecoStar applications that was born from practical experience. All parameters for character recognition and image processing are convenient to configure and can be tested interactively using individual images or batches of images. The results are graphically displayed and synchronised with the document image. Intermediary steps in image processing can be analysed individually in a synchronous before/after display. A convenient do/undo functionality permits comparison of results achieved using different parameter definitions.

Migration of RecoStar 3.x applications
Not all functions of RecoStar 3.0 applications are currently supported at the .Net interface without modifications. The new RecoStar, however, provides alternatives to most of these functions, which can be used for migrating the application.
The RecoStar Design Studio provides for this purpose a convenient migration support. Synchronised with the display of the migrated project, the developer is provided with precise information about any restrictions that may have occurred and their cause, and this makes for a very efficient migration and testing process.

Alternatively, it is possible to operate an existing RecoStar 3.0 project unchanged at the .Net interface. The recognition results can then be analysed with the new results parser. In this scenario, the project is maintained using the well-known 3.0 IFG. This approach is especially useful when migrating an application park in the field step by step.

Deployment
Merge modules simplify the integration of components in another application. That is why Open Text Capture Recognition Engine 4.0 contains a merge module for installing the runtime engine. This can be used for system integrations on the basis of both Open Text Capture Recognition Engine 3.0 interfaces and the .NET interface.

New form type recognition
Many forms lack a form key, and instead have to be identified using graphic objects in the image. A new and very easy to use, self-learning form type recognition feature is integrated in RecoStar Professional 4.0. Examples of the forms requiring identification are stored for training. In the learning phase, Open Text Capture Recognition Engine then analyses which line structures and other features are characteristic for the document. During production, the documents to be classified are identified according to their similarity.

Reading check boxes
RecoStar 4.0 Professional introduces a new field, ‘Reading check boxes’. It enables straightforward definition of boxes and box groups. In addition to purely evaluating the degree of blackening in the check box, use is made here of our proven classification and image processing technologies – all of which lead to very high recognition rates.

Improved recognition performance and speed
Version 4.0 marks yet another increase in the Open Text Capture Recognition Engine recognition performance. The algorithmic improvements benefit in particular the recognition of typed text and the highly precise reading of (German) numerical amounts.

Reading speed for large typed documents is of considerable importance for many applications. Considerable improvements have been achieved here too.