The Solution
Image-X provided a distributed scanning,
processing, indexing and retrieval system to take care of
the forms processing and document management
problem. The document imaging system is integrated with
the existing Health Management application which is running
on an HP 3000 computer and provides a comprehensive health
management system for all branches. The components of the
system include one scanning station, one optical character
recognition and processing station, and one quality assurance
station. The OCR-processed text is stored
on an HP 3000 based health management application. The scanned
documents are stored on a Hewlett-Packard optical jukebox
with a capacity of 40 GB, connected to an HP 9000 computer.
This provides an online capacity of nearly one million documents.
The forms are first batch scanned into the document
management system using Fujitsu scanners. These
scanned forms are at first stored on the local hard disk,
to allow the scanning operator to rescan the documents if
the quality is not good. After a batch of documents has
been scanned, it's automatically forwarded to the optical
character recognition (OCR) server. The
OCR server processes the forms and sends the results to
the quality assurance (QA) station for verification. The
QA server is the only place where human interaction is required.
The QA operator views the original images and the OCR
process results on the QA station. The fields that are not
verified are flagged by the software and displayed along
with the original text. After the OCR-processed
text has passed the quality assurance test, it is automatically
sent back to the OCR server. The OCR
server then transfers the scanned images to the optical
jukebox attached to the image server. The recognized fields
from the forms are used to update the database application
running on an HP 3000.
Image-X's document imaging system provides
document imaging and management solutions tailored to the
needs of the customer. The components of the imaging system
are modular and can be configured to provide a highly customized
solution. The system is available on PCs, UNIX, OS/2 and
Macintosh machines and can be integrated with any database
application running on any kind of hardware platform. In
addition to superior document management for millions of
documents which results in fast search and retrieval speed,
the unique indexing architecture automates the process to
eliminate manual key-entry. The neural network based OCR
allows automatic deskewing, forms recognition and line removal,
and accurate OCR-processing which results in significant
savings.
|