[2539] in comp.os.os2.announce archive
PR: Solution Technology Launches True Omni-forms Processing
daemon@ATHENA.MIT.EDU (Leon D. Zetekoff)
Sun Feb 23 21:27:24 1997
To: os2ann.DISCUSS@bloom-picayune.MIT.EDU
Date: 21 Feb 1997 17:54:05 GMT
From: solution@gate.net (Leon D. Zetekoff)
Reply-To: solution@gate.net
Reply-to: solution@gate.net
---------------------------------------------------------------------
Date: February 20, 1997
Solution Technology, Inc. 1101 South Rogers Circle, Suite 14
Boca Raton, FL 33487
Contact: Leon D. Zetekoff, N.C.E. (solution@gate.net)
Voice: 561-241-3210 Fax: 561-997-6518
E-mail: solution@gate.net Web Site: http://www.gate.net/~stidev
Solution Technology Launches True Omni-forms Processing
Boca Raton, Florida--After quietly fielding initial systems to selected
corporate and government clients, Solution Technology is now moving its
modular Genie , omni-forms processing system into general release. These Genie
systems accommodate different forms processing volumes, ranging from a single
standalone system to multi-station, networked, parallel processing workflow
environment systems. Typical client volumes range from a few hundred per day
to 26,000 forms per day. Genie can be modularly expanded to accommodate nearly
any volume.
Genie scans and imports images, recognizes forms, performs OCR/ICR, barcode
and mark sense, does edit/repair, and exports indexed images and data to
existing imaging and data base systems. Genie allows customers to build high
performance, cost effective, omni-forms processing systems that do not require
end-user or custom programming.
Form Recognition uses STi's unique omni-form "Form Fingerprinting" technology.
Form fingerprints are generated automatically from a reference image. Images
are corrected for distortions in both axises for precise registration. The
registered image is converted into a compacted spatial abstraction. Unknown
forms are similarly fingerprinted and then matched against known form
fingerprints. The STi form recognition processors can discriminate forms with
as much as 25% damage and a 5% scale difference between the reference and the
image to be processed. Rejected unknown images are shunted aside for manual
processing. The recognized candidates are organized into a list ordered by
match criteria. Optional secondary recognition processors examine symbols,
titles, barcodes and other form identifiers to discriminate between forms with
otherwise similar fingerprints.
New form templates are easily added by scanning a reference form and selecting
learn. You designate the areas of interest with a mouse, assign a field
processor (Barcode, OCR/ICR, etc.) and name the field. Additional fields can
be easily added or deleted at any time. There is no limit to the number of
fields per form as well as no practical limits on the number of forms.
Edit repair stations correct and dispose of forms quickly. The editor
displays original image and field data to be corrected to minimize operator
fatigue. Fields without errors are auto-skipped, stopping on the next field
to be corrected.
Genie's form reader servers support multiple OCR/ICR engines on the same form.
Current engines include plug-ins from Nestor, Mitek, as well as STi's own
engine. Other plug-ins include an "All Angle Barcode Anywhere ", field
barcodes, patch codes, mark sense, signature detection and extraction, name
and address block decoding, tabular data processing and multi-field data
roll-up.
Genie is available on OS/2 with NT currently scheduled for release by 2Q97.
TCP-IP and NetBIOS communication protocols are supported. STi's native 32 bit
TWAIN for OS/2 and TWAIN for NT support desktop scanners from Hewlett-Packard
through Kodak 900 at their rated speeds. Most industry and corporate image
file formats are also supported including TIFF (i.e. CCITT-G3 and CCITT-G4),
MO:DCA/IOCA, PSEG, GIF, PCX, DCX and many fax file formats. Genie supports
direct export of indexed images to imaging systems (i.e. IBM ImagePlus) as
well as user scripted ASCII transaction record files.
=====================================================================
| c.o.o.announce submissions: POST [or email comp-os-os2-announce@uunet.uu.net]
| Please address other correspondence to lfirrantello@bix.com .
| Please see: http://www.bix.com/pub/os2ann/pindex.htm for posting guidelines
