Recoflex

Cardiff TELEform’s RecoFlex recognition engine recognizes all of your data, including hand print (ICR), machine print (OCR), bar code, check box (OMR) and signature fields.

RecoFlex offers sophisticated algorithms that improve TELEform’s recognition for both machine and handprint to increase the level of automation for a TELEform system. The voting algorithms are dynamically invoked based on the specified field attributes such as the font size, font type and language. The engine mixes are also dynamically selected for the voting algorithm based on field attributes.

Additional verification options include the ability to select flexible processing options with processing speed and accuracy adjustments.

For most applications, TELEform Reader will correctly convert information locked on paper at an accuracy rate over 90%. RecoFlex is the ultimate neural-based recognition system that combines multiple recognition engines and applies sophisticated ‘voting’ to deliver more accurate results. TELEform provides point-and-click settings that allow for precise control over recognition settings, including tuning for alpha/numeric, special/International character sets, database/table comparisons, and even signature/initial fields. These settings clearly increase accuracy and throughput for your application.

RecoFlex improves ICR accuracy with the addition of the following powerful engine: KADMOS 3.5 – Highest rated German ICR engine.

The result is improved data accuracy and less time spent in verification. As employees spend less time in the verification process, organizations can gain significant increases in throughput and employee productivity without sacrificing accuracy or efficiency. The bottom line: higher revenues and greater profitability.

Enhancements to the RecoFlex engine offer ICR recognition improvements in addition to those OCR improvements already introduced with v8.0. TELEform v8.2’s higher ICR recognition accuracy rate further increases the level of automation for a TELEform system. Updated voting algorithms are dynamically invoked based on the specified field attributes to increase processing speed.

Supported OCR sets include Omnifont, dot matrix, OCR A, OCR B, Farrington
Supported OMR sets include check marks, bubbles, circle responses, underlines, brackets
Supported barcode types include PDF-417, Code 39, 93, 128, Interleaved 2/5, UPC, EAN, EAN/UCC 128, Codabar, ISBN Add-on 2 and 5, Patch code (recognition only), Industrial, Datalogic, Matrix, IATA 2/5 (recognition only)
Supported language sets include Polish, Turkish, Icelandic, Hungarian, Slovenian, Lithuanian, Greek, Dutch, Czech, Russian, Bulgarian, Slovak, French, Ukrainian, Belgian, German