Table of Contents
Why use FlexiCapture Technology?
The term “ABBYY FlexiCapture” is used in slightly different contexts:
- As a name for a product line of professional applications (Standalone, Distributed) to process forms, classify documents and extract data. The structured information then can be used in business processes.
Further details on the FlexiCapture products and solutions can be found here: English, German, French.
- FlexiCapture Engine provides the core technologies for developers as toolkit.
- FlexiLayouts and FlexiLayout Studio are the heart of the technology to develop and define document definitions and the extraction logic to make the magic happen.
Below some fundamental differences between custom rules to extract data from raw text and FlexiLayout technology.
Parsing Fulltext OCR results
- Simple data extraction from “plain OCR text” files often uses regular expressions .
- But in structured documents the layout contains very valuable information. It is not just logos, image and text areas, but also lines and gaps between different elements which structure content on a document.
- ABBYY OCR technology is able to analyze the layout of PDFs and images. Developers get access to the layout structure and the coordinates of certain regions:
- The information is accessible via:
- FineReader Engine layout object - “live” access during document processing
- the ABBYY XML - “offline” access to parse the information
Both approaches mentioned above work, but the ABBYY FlexiCapture Platform offers further advantages …
- “Simple” full text parsing is not often explicit
- –> FlexiCapture Technology goes far beyond
- Writing „hard coded“ extraction logic can be complicated and the written code is hard to maintain
- –> FlexiCapture Technology is productised technology and comes with training & support
- → there are certified experts on the market
- Dealing with text and layout coordinates in pure code is very abstract
- –> FlexiCapture Technology comes with
- Visual Development Tools
- Logic to deal with graphical elements
- Logic to use the relation between objects
- Automatic Hypothesis Analysis – also allowing optional elements
The illustration shows the differences between the document analysis of an invoice for fullt-ext OCR and the same document processed with an own FlexiLayout on top of the pure block analysis.
FlexiCapture Technology also provides:
- Multi-page document support is built in
- Complex document structures with repeatable elements, e.g.
insurance contracts with a form for each family member, embedded in a standard set of other pages
- Scripting support to define what ever you need
Flexible Data Extraction in Mobile Scenarios
- If data should be extracted on mobile devices - this is even harder to realize, because:
- The image quality is often not as good as the scanned documents
- OCR SDKs for mobile devices do not have the same layout analysis capabilities as “big” PC/Server SDKs
- Scanners have a fixed geometry.
- So the information on the same type of documents will have more or less at the same coordinates.
- When a camera is used to capture the documents, the distance between the document and the sensor are most likely never the same.
- This results in a different reproduction scale and there fore the information will be located at different coordinates → a simple approach with coordinate look up will fail.
- FlexiLayout technology does not use an absolute set of coordinates so it can deal also with mobile images.