Theale, UK (9th October 2012) - ABBYY®, a leading provider of document recognition, data capture and linguistic software and professional services, today announced ABBYY FlexiCapture Engine 10, the next generation data capture and document classification software development kit (SDK). The new toolkit combines ABBYY’s best-in-class optical character recognition (OCR), imaging and data capture technologies with exceptional ease of use through the integration of innovative auto-learning technologies, enabling developers to create and train data capture and classification applications by simply pointing, and clicking.

FlexiCapture Engine 10 enables developers, corporations and service providers to integrate award-wining data capture technologies for document processing and data extraction from invoices, contracts, forms, and more. It also delivers accurate document classification, mobile data capture and export for backend processing.

FlexiCapture Engine 10 enables application developers to create solutions that benefit business processes by:

Solving the problem of capturing paper documents and inputting them into the corporate ECM by automatically extracting metadata from scanned documents, enabling users to better classify and name documents in their ECM or DMS systems.

Enhancing business solutions with easily trainable document classification and data capture capabilities, improving efficiency and increasing ease of use.

Empowering corporate document management infrastructure with mobile data capture capabilities. ABBYY‘s unique FlexiLayout data capture technology is reinforced with mobile photo pre-processing, scalable multi-core processing, prebuilt mobile data capture web-service and demo samples for quick and easy integration.

“Our new FlexiCapture Engine SDK has made a huge leap forward removing the major adoption barrier and simplifying document definition creation,” said Andrey Isaev, Director of SDK Products Department at ABBYY. “Its new auto-learning functionality not only enhances document processing capabilities, but shortens learning curves and speeds application time to market.”

New features of ABBYY FlexiCapture Engine 10 include:

Auto-Learning and Classification Technology
New auto-learning tool enables developer solutions to automatically learn and classify new document types following initial interaction, and self-identify relevant fields for data extraction.

Enhanced Image Pre-Processing
Improved ABBYY Camera OCR technology enables FlexiCapture Engine 10 to achieve better results from photos through enhanced removal of motion blurs, noise in color photos, and geometrical distortion from an image. In addition, FlexiCapture Engine 10 delivers new imaging technologies that prevent the loss of information and result in higher quality text retention such as removing security paper guilloche background, suppressing color objects in official documents and creating custom image filters to extract maximum information from hardly readable documents.

Increased OCR Accuracy for Asian Languages
New search mode improves recognition accuracy of text lines in Chinese, Japanese and Korean languages. In addition, recognition of 2D barcodes including QR Codes, Aztec, and Data Matrix is now supported.

Additional Customization Options for Better Recognition Results
ABBYY FlexiCapture Engine 10 enables new verification capabilities via special API which provides access to different hypotheses of character or word recognition with corresponding confidence values. This API can be used to check recognition results using databases and algorithms, enabling less time to be spent on manual verification and improving the quality of extracted data.

Updated Developer Samples in Code Sample Library
FlexiCapture Engine now features two more code samples including an auto-learning technology demo tool and an image preprocessing sample. The first sample code allows classifier creation and document definition. The second sample includes a set of image preprocessing tools that allows developers to watch how tools influence recognition quality, and select the appropriate image pre-processing technology for the particular document type.