Compare and Contrast OCR and Forms Recognition Technologies Peter Lang and Scott Hamilton
Agenda Capture in ECM Choices, choices Product Overviews - Peter ABBYY FlexiCapture TeleForm Product Overviews Scott Kofax KTM Oracle Forms Recognition Product Matrix Final Comments and Questions
What is our main focus today? Forms Recognition Separation Classification Data Extraction from Scanned Images Handwriting Machine Print OMR Bar Code
In the multi-dimensional ECM realm, Capture is the on-ramp
Where is Capture going? From AIIM, 2010 Research Paper The strongest driver for scanning and capture is improved searchability and knowledge sharing across the business, followed by productivity improvements, reduced office costs and better customer service. respondents are posting very positive preferences for both centralized and distributed capture in-house, and are making big moves towards automated data recognition and document auto-classification..
When Fruit Hangs Low and Heavy Harvest! 66% have formal scan-to-archive process 47% utilize workflow Only 16% scan/extract data to a process Only 50% use automatic recognition of metadata for indexing. Only 14% use capture and BPM across multiple processes/depts Continued from AIIM 2010 Drivers and Experiences of Content-Driven Processes
Getting On Board Or Ramping It Up 39% reach positive payback in 12 months 60% within 18 months. Automatic document classification shows a particularly high return for the 19% of respondents utilizing it. Continued from AIIM 2010 Drivers and Experiences of Content-Driven Processes
Choices, choices. If 50% use automatic recognition of metadata for indexing Huge potential for time and cost savings. Many ways to bridge this gap If automatic document classification shows a high returns for the 19% of respondents utilizing it. Many options will allow you to join this group of satisfied respondents
Esoteric Ancient Wisdom Break? Paul Reps, Zen Flesh Zen Bones
What s best for you? Today we ll look at Cardiff TeleForm ABBYY FlexiCapture Oracle KTM Oracle Forms Recognition All of these products can take you to the next level of accelerated processing, cost savings, and, along with other offerings, are all enticing technologies to take advantage of.
ABBYY FlexiCapture ABBYY was founded in 1989 by David Yang. By 2009, the company had over 900 employees in nine offices in: Germany (Munich), the UK (Bracknell), the USA (Milpitas, CA), Japan (Tokyo) Taiwan (Taipei) Russia (Moscow and St. Petersburg) Ukraine (Kiev) Cyprus
ABBYY Products FineReader Convert doc images PDFs into editable/searchable files FlexiCapture Auto processing of multiple doc types in a single stream PDF Transformer PDF conversion and creation
ABBYY FlexiCapture ABBYY FlexiCapture is available in two configurations: Standalone Installation on one machine One operator processing /verifying forms Typical volume: 1,000 3,000 forms per day.
ABBYY FlexiCapture ABBYY FlexiCapture is available in two configurations: Distributed Flexible system architecture Multiple workstations Process 10,000 or more forms/day.
ABBYY FlexiCapture Modules Processing Server and Processing Stations Automatically execute all resource-intensive operations Image import recognition, classification, data extraction, validation rules execution, and export. Administration and Monitoring Console Manage processing and track statistics.
ABBYY FlexiCapture Modules Application Server Handles all processing tasks Automatically queues/distributes among the stations FlexiLayout Studio Semi-structured and unstructured documents Example: Invoices, contracts ABBYY Project Setup Station Instruments to test, fine tune and adjust projects
ABBYY FlexiCapture
ABBYY FlexiCapture While being easy to work with, FlexiCapture technology allows users to fine-tune data capture logic Provides administrators with direct access/control over generated document descriptions Helps to manually change descriptions if necessary Achieves best possible accuracy in locating/recognizing data fields on a document
ABBYY FlexiCapture Manual tasks occur during Quality Control and Correction. A batch will enter Correction only if data entry fields in the batch were not evaluated with sufficient confidence. A verification processor would then check the suspect data entry fields and make corrections as needed.
ABBYY FlexiCapture Demonstrations
Autonomy TeleForm Designed in 1991 Built to capture form data from faxed forms Currently handles fax, paper, and electronic forms Current Version: 10.5.2 Acquisitions & Rebranding Cardiff acquired by Verity, Inc. (2004) Verity Teleform Verity acquired by Autonomy (2005) Autonomy Cardiff Autonomy acquired by HP (2011) Despite all of this: just go to www.cardiff.com
Autonomy TeleForm Configurations Desktop Standalone server/workstation Process from 500-1500 forms/day. Workgroup Can include up to 20 modules) Enterprise No restrictions on the number of module/workstation licenses.
Autonomy TeleForm Modules Designer Module Design new or existing forms Look, layout, field attributes, validations and data export requirements all defined in this module. Scan Station Loads job specific settings Form types Batch specific scanner settings.
Autonomy TeleForm Reader Module Image de-skew and clean-up Form identification OCR/ICR/OMR etc. All first pass validations. Verifier Module Reviews failed validations and business rules Reviews characters that fall below defined confidence thresholds
Autonomy TeleForm
Autonomy TeleForm Manual tasks occur during Quality Control and Correction. During Quality Control operator tasks: ensure that the batch contained valid files classify batch items if necessary collect index information improve the quality of images in the batch flag potential items for later review.
Autonomy TeleForm Demonstrations Scan Station, Reader, Verifier and Designer, Traditional Form Designer, Existing Form Verifier, Handwriting Search When Handwriting Is Your Only Option for a relevant blog by Pete.
Capture Matrix 1-9, 9 Most Desirable Setup Speed Structured Setup Speed Semi-structured ABBYY FC Cardiff TF Kofax KTM Oracle FR 7 9 7 6-7 6 6-7 Handwriting 9 9 4 Semi-Structured Form / Data Recog 9 7 9 Classification 8 7 7 Easy To Learn 6 7 5 Multi-Languages 9 7 9 4 Quality Control 7 8 7
Questions?