ABBYY unveils FineReader Engine 11 OCR SDK
ABBYY has launched FineReader Engine 11, the newest edition of its optical character recognition (OCR) software development kit (SDK) used by many global OCR vendors.
ABBYY FineReader Engine 11 introduces various enhancements including an API for automatic document classification; a new business card recognition capability; extended PDF export, enhanced image preprocessing, improved Word and Excel export; and native 64 bit and easier Java integration.
It offers high accuracy OCR for more than 200 languages including improved recognition of Arabic and is available for the Windows, Linux and Mac platforms.
“ABBYY has long been committed to pushing the envelope on OCR performance and accuracy, and FineReader Engine 11 represents the culmination of this effort,” stated Andrey Isaev, director of Technology Products Department at ABBYY.
“The latest generation of our SDK brings together numerous enhancements that software vendors have been asking for to satisfy the challenges their customers are facing. It also raises the bar on accuracy and language support, enabling development teams to not only save time and money themselves, but pass it onto their own customers, making them more competitive.”
New features and enhancements in ABBYY FineReader Engine 11 include:
Automatic Document Classification — via combination of content- and image-based classification technologies the application can detect what document types are being processed, so developers can easily sort a batch of mixed documents automatically;
Business Card Recognition — full-feature business card recognition technology is now accessible through an API, enabling developers to incorporate data extraction and export into their applications;
Extended PDF Export — for further support and compliance with archiving standards, the FineReader Engine now provides export to PDF/A-2, a format which enables saving to smaller file size that leads to reduced storage space for long term archiving, and PDF/A-3, which enables inclusion of binary format such as XML or Microsoft Word. Additional PDF processing enhancements include improved MRC compression, PDF bookmark preservation and increased export speed;
Improved Arabic, Chinese, Japanese, and Korean OCR — this version offers Arabic OCR that is twice as accurate and three times faster than it was in FineReader Engine 10, as well as improved speed and accuracy for Japanese, Korean and Chinese traditional and simplified documents.
Improved Word and Excel Export — improved layout retention and font detection in Word and Excel documents during export.
Native 64-bit and Simplified Java Integration — improved 64-bit support allows developers to take full advantage of 64-bit hardware; and new implementation-ready Java classes eliminate the need for significant coding to create wrappers for Java applications.
Enhanced Image Preprocessing Tools — improved algorithms for geometrical distortions correction, new auto-cropping and background lightening features, and better ISO noise removal ensure more accurate recognition results, especially when documents are captured with cameras in mobile devices and processed in an OCR back-end system.