ABBYY 3A Helen Sapronova Catherine Moskalenko 2 2 Catherine Moskalenko Sales amp Marketing Specialist Dedicated webinar Host in ABBYY 3A Helen Sapronova Product Manager SDK Regional Manager ID: 769926
Download Presentation The PPT/PDF document "ABBYY 3A Helen Sapronova, Catherine Mosk..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
ABBYY 3A Helen Sapronova, Catherine Moskalenko
2 2 Catherine Moskalenko, Sales & Marketing Specialist Dedicated webinar Host in ABBYY 3A. Helen Sapronova,Product Manager, SDK,Regional Manager, China & South Korea One of the most experienced and respected professionals in ABBYY 3A, brilliant technical specialist.
Agenda What is new in ABBYY FineReader EngineImprovements in architectureDocument ClassificationBCROCR improvements Other improvementsUpdate of Sales & Marketing Materials 3
why… OCR? 4
OCR business is declining? 70% of organizations perform the initial scanning of documents, but only 13% are extracting the data via OCR and processing it Nearly half (47%) of them have made only 5% progress towards processes that could be paper-free, 18% haven't even started yet1/3 are processing electronic documents, forms and PDFs separately from scanned paper. 20% print them out - including 13% who print them out and then scan them back into the capture systemDriving paper out of the process would improve speed of response by a factor of 4.6x and improve the productivity up to 35%5 2/3 of enterprises adopting paper-free processes report a payback within 18 months, 1/2 see payback in a single 12-month budgeting period *According to ©AIIM 2013, www.aiim.org
ABBYY FineReader Engine is … … an OCR SDK that gives developers, integrators and BPOs the tools they require to integrate text recognition technologies into their applicationsIt allows companies to provide their customers with the best products and services, thus leading the business to a new level of success6
What is new? 7
Key New Features 8
Architecture improvements ABBYY FineReader Engine 11
FineReader Engine 11 is now available for Windows, Linux and OS X platforms ABBYY FineReader Engine 11 is immediately available on the Windows, Linux and OS X platforms Linux and OS X versions have subsets of full feature list New! 10
64-bit native support Native and neutral 64-bit applications development is possible now ! Avoid “fragile” code in your applications C++ DLLs now could be linked in x64 applications directly without using COM proxyNeutral .Net/Java interops that can be used for building .Net/Java project for “any CPU” included New!11
Document classification ABBYY FineReader Engine 11
Document Classification Automatic, universal, easy to integrate Allows to assign a document to some class on the base of its content and/or appearance New! 13
Classification usage scenarios [Archiving] Sorting documents by type for electronic archive creation [Mailrooms automation] According to detected document class further routing can be initiated [Batch processing] Document separation[BPO] Pre-sort documents for further processing[Banking/Insurance] Verify document set completeness applied to get loan/ insurance payout[OEM] Smart MFP/scanner interfaces suggesting typical actions for each document class 14
Advantages of Classification in FineReader Engine 11 Automatic No templates required. Can be trained easily by non-technical end-users UniversalFits for all types of documentsEasy to integrate One page of code required forbasic scenario15
Classification profiles Maximum accuracy: Based on full-text OCR analysis key words are detected automatically during the trainingMaximum speed: Based on Title text OCR & image pattern16
How to start classification? 17
Classification cross-positioning 18 ESDK FRE FCEFCInput ImageImageImageImage Layout-based - + + + Content-based+ +++Automatic Learning+ +++User-defined classification rules--+1+1 available through FlexiLayout Studio interfaceGeneral difference between classification in ABBYY products is in products themselves: they are intended for different tasks
Classification Demo 19
Business card recognition ABBYY FineReader Engine 11
Business Card Recognition 27 languages supported API provides access to the following types of extracted data:Personal name Company name Position in the company Company address Phone number Fax Mobile phone number E-mail Web site Auto-splitting of multiple business cards scanned on one page New!21
Business Card Recognition Recognized data can be saved in vCard, CSV or XMLIdeal for software or online service bundled with scanners/MFPs22
PDF support improvements ABBYY FineReader Engine 11
Faster PDF Export Export to PDF now is up to 12% faster than in v10 24
Reduced PDF output size Output PDF MRC file is up to 50% smaller Compression is more efficient for documents containing text than for documents containing many images + text25
PDF Bookmarks support In “PDF -> searchable PDF” scenario there are source PDFs that contain bookmarks. Up to V11 these bookmarks got lost during export. Now bookmarks can be retained in the output PDF file *Available in Windows and Linux versions New! 26
Export to PDF/A-2 PDF/A-2 is an extension to the existing archiving standard PDF/A-1 PDF/A-2 introduces a number of useful features, such as JPEG2000 compression and new conformance level PDF/A-2u especially for Unicode*Available in Windows and Linux versions New!27
Other PDF improvements Special processing mode for text-based PDFs FineReader Engine 11 includes API to omit OCR of text-based PDFs, or re-use text information from them:In the scenario “image PDF -> searchable PDF” one may do not perform processing of the text-based PDF at all, but simply copy original file to the output folderIn the scenarios when text-based PDF is converted to some other format, one can select whether the text information from the file should be used during processing*Available in Windows and Linux versions 28
OCR improvements ABBYY FineReader Engine 11
More accurate Arabic OCR In FineReader Engine 11 the number of incorrectly recognized words for Arabic OCR is 2 times less compared to v10! 30
Faster Arabic OCR At the same time with great accuracy improvements, recognition speed for Arabic texts in FREngine 11 became approximately 3 times faster compared to v10 31
Faster Chinese/Japanese/Korean OCR Chinese (Simplified) is 2.5 times fasterChinese (Traditional) is 4 times fasterKorean is almost 3 times fasterJapanese is 2.5 times faster 32*Fast mode, 4 cores CPU
Other improvements ABBYY FineReader Engine 11
Auto-splitting of double-page spread Better appearance of output document (page-by-page) Higher effectiveness of image preprocessing (curved lines correction, scanning shadows removal) Previous version of FineReader Engine can find position of a double-page spread. The split itself should be performed manually. New! 34
New barcode types autodetection Maxicode barcode is used for tracking and managing the shipment of packages (i.e. by UPS company) USPS 4CB barcode - in FRE11 Maintenance release35 New!
New font management API Font management is much easier and provides a variety of predefined font filters which save developer from manual font specifying: default set used by ABBYY FineReader a set for European languagesa set for Chinese languagea set for Japanese languagea set for Korean languagea set for Arabic languagea set for Hebrew language a set for Thai languagea set for Armenian language36New!
Scanning API improvements Asynchronous scanning allows to run recognition of scanned pages before scanning of all pages is finished Extended access to scan settings, including access to scan source capabilities Filtration of scan sources by available user interfaces (FineReader, scanner UI, none) or scan API types (TWAIN, WIA) All limitations for implementing service-like scanning has been removed (writing log file can be canceled, scanning does not require Registry access)An ability to specify compression type of scanned images*Available in Windows version only37
New code samples & updated Help file Classification Business card recognition FineReader Engines PoolScanning BatchProcessor38 New!
39 How to: Effectively use resources of a high performance computer FRDocument API for parallel processing of multi-page documents. Preprocessing, analysis and recognition are performed in parallel; synthesis and export are performed sequentially in the main process BatchProcessor API to process many one-page documents . Preprocessing, analysis and recognition are performed in parallel; synthesis and export are performed sequentially in the main process while new pages are being analyzed and recognized in other processes To perform full processing of many one-page documents in parallel use a pool of Engines loaded out-of-process by means of COM
Demo 40
41 updated sales & marketing
Price lists and Marketing materials Basic prices remain similar to previous version New add-on modules:- Arabic OCR: 30%- Document Classification: 20%- BCR: can be included into Professional Runtime licenses purchased with SMUA Updated SDK License Agreement Template – ask your managerUpdated materials will be available on Partner Portal shortly:- Price lists- Product Brochure- Product Presentation42
Powerful OCR Engine to Win the Business Race QUESTIONS?