A

AccuID

An OCR for AnyDoc method for identifying master form templates. This method works at the form family level to build an identification table based on the unique topology of each master form template in the form family and compare it to incoming data images. Then AccuID sorts and identifies data images from the correct master form template for processing.

AccuZip

OCR for AnyDoc accesses the AccuZip database to validate addresses in the United States of America. It is capable of processing address information from all 50 states, and from virtually all U.S. territories and military installations abroad. AccuZip is available as an add-on feature for OCR for AnyDoc.

Address Extraction

The Address Extraction feature of OCR for AnyDoc allows you to extract U.S. and Canadian address data off your documents and define how the data will display in your output.

Address Extraction gives you the option to validate address data using AccuZip.

AnyApp Technology

AnyApp locates data on template-resistant forms—regardless of where it resides on the document—by searching for defined data labels, such as “amount due” and “invoice number”; data format, such as dd/mm/yyyy; data type, such as alpha, numeric or both; and/or location, such as “just look in the top half of the document for this data.” It then remembers where it found the data when the document type is processed again. AnyApp is the technology behind the AnyDoc Software solutions AnyDocEOB and AnyDocINVOICE.

Attachments

In OCR for AnyDoc, attachments are document images to be archived along with the processed document.

Audit Phase

During the Audit phase of OCR for AnyDoc, the Auditor module allows you to review the work of your verification operators. It allows you to audit specific tasks, do random checks, or review all of an operator’s work.

AutoFlow

An automated method used to check available workstations for batch processing jobs that need to be completed.

B

Bar Code Zone

A master form template zone that defines the location of bar code data in the document image. OCR for AnyDoc recognizes these bar codes and converts them into alphanumeric data.

Batch Separator Page

Printed for a form family to provide a fast and accurate method of entering control information for a given batch of forms.

BCR

Bar Code Recognition. The process of reading and extracting data from bar codes on a document. See also Bar Code Zone.

C

Caere

An OCR engine developed by Caere, Incorporated (now Nuance Communications, Incorporated).

Capture

Document capture is the method of obtaining the document image (either from scanning or importing) from which OCR for AnyDoc will extract data. Data capture is the extraction of this data, which can be used in a database or back-end system.

Character Constraint Boxes

With character restraint boxes, you restrict the amount of data that can be entered on a form by providing a specific number of boxes to be filled in by the user.

CMS 1500

The standard form from the Health Care Finance Administration, designated for submitting healthcare claims to insurance companies. Previously known as the HCFA 1500.

Commit Phase

The last phase of OCR for AnyDoc batch processing prior to data output, where output files (e.g., TXT, GTO, XML, PDF) and archive images are written to the appropriate directories.

Conditional Procedure

A user-designed routine that features advanced character searches, recognition and replacements. A conditional procedure retains or filters data, based on the specific condition.

D

Data Capture

The ability to capture digital data off scanned paper document images. This data then can be transmitted to a financial or back-end system for entry into an ODBC-compliant database.

Date Extraction

A feature in OCR for AnyDoc that automatically converts extracted date information from multiple input formats into a user-defined output format.

Delimiters

Special characters that separate data fields and/or records so the data can be parsed from the file by a program or a script.

Distributed Capture

A means by which organizations can scan documents remotely, either from branch offices around the world or simply downstairs. The scanned images are then transmitted via a secure Internet connection for data capture processing at a centralized location, such as corporate headquarters.

Document Set

A group of related documents that need to be processed together as a batch.

E

EDI

Electronic Data Interchange. The transfer of data from one business to another over a network.  

Endorser

A mechanism found on some scanners that print an incremental number on an image, which facilitates document indexing.

EOB

Explanation of Benefits. A statement from a healthcare provider that itemizes how benefits were approved or denied for a claim.

External Table

A database table connected through an ODBC link.

Extract Process

An OCR for AnyDoc batch control process that de-skews the image, performs form removal functions, enhances images, regenerates characters, applies pre- and post-processing rules that have been set, etc.

F

Form Family

One or more master form templates grouped together for batch processing.  Examples of form families include a batch of invoices that must be batch-balanced and a mortgage folder that has pages to be processed, containing information used to index other pages in the folder within an image retrieval system.

Form families can do the following:

• Archive images
• Batch balance controls
• Create a header record
• Enhance form identification
• Name directory structure
• Perform batch verification

H

HCFA 1500

See CMS 1500.

High Speed Verification

An optional verification phase in OCR for AnyDoc batch processing in which operators see only the image’s questionable characters. Verification is “high speed” when the operators can correct at once all questionable characters in a batch, rather than tabbing to each questionable character on a data image.

I

ICR

Intelligent Character Recognition. The process of converting handwritten characters into ASCII text through the use of a recognition engine.

Identify Phase

The phase of OCR for AnyDoc processing, during which AutoID automatically identifies each document type.  AutoID uses static elements on each document, such as barcodes, literals or graphics, to identify the document.

Image Registration

The use of an image on a document (such as a square or a triangle), both contiguous and containable, as a registration point to help OCR for AnyDoc auto ID a document type.

Import Phase

The process of bringing images into OCR for AnyDoc with or without the use of a scanner. Import is a batch processing phase in OCR for AnyDoc.

Indexing

A means of electronically identifying a scanned document image for archival and retrieval purposes.

Intelligent Extraction

On an OCR for AnyDoc template, Intelligent Extraction recognizes a date, an address, or a currency type and converts that data zone into a user-specified format.
For example, all dates can be output into the format MM/DD/YYYY, no matter how the date is written on the document.

Inverted Text

The placement of white text on a black background in a scanned document image. The text and background must be inverted for OCR for AnyDoc to read the text.

J

Job Queue Directory

A temporary network location where OCR for AnyDoc stores its processing files. These files identify the status of each job/image/page.

Job Manager

In OCR for AnyDoc and AnyDocCAPTUREit, Job Manager is a server component, typically installed on a network server, that is used to facilitate automated remote data capture.

K

Key

A key (or primary key) is a field that uses a number or character sequence unique to each record in a table (e.g., social security number) for identification purposes.

Key-from-Image

The process by which data entry operators view and key data off electronic, rather than paper, versions of documents. The key-from-image approach to data entry is approximately 10% more efficient than traditional data entry methods.
In OCR for AnyDoc, the key-from-image verification GUI is the default verification method. With the program’s rope and expand capabilities, however, operators key significantly less data.

In BROKERit, key-from-image activities occur in the query module.

L

Literal

Synonymous with text. A literal can be machine print (OCR) or handprint (ICR). Static literals on a document can help OCR for AnyDoc with registration points and to identify a form type.

Lookup Table

A table that OCR for AnyDoc accesses to validate specific data residing on a processed document.
For example, OCR for AnyDoc can access a P.O. number table to validate the vendor associated with the P.O. number on an invoice.

M

Manual Indexing

In AnyDoc®BROKERit™: the assignment of areas in a particular document to a particular field in a document or data table.

Mark Sense

Data confined to one or more selections in a series, as in a survey. The data is selected by checking a box or filling in a bubble.

For example, a survey may include gender information. A respondent fills in the bubble next to ‘M’ or ‘F’ on the survey to indicate his or her gender. OCR for AnyDoc seeks that mark sense zone for the data and extracts the selected response for that question on the form, based on the pixilation present in the selected bubble.

Mark Sense Mark

A mark on a form identifying a selection of mark sense data. The mark consists of the presence of pixels (such as a check mark, a filled-in box, a signature, etc.). The recognition engine searches for the presence (a “hit”) or absence (a “miss”) of a mark.

Master Form Template

Scanned or imported document images used to define the zones and parameters for processing data from structured documents of the same type.

N

Noise Filtering

Removes particles (black dots representing noise) from the document image.

Note Zone

Note zones define areas of the form containing data that are not processed by an OCR or ICR engine. OCR for AnyDoc prompts the operator to enter the data during verification. A note zone is useful for obtaining data such as signatures or other unconstrained handprint.

O

Omit Zone

With omit zones, you define the areas of a document to be ignored during OCR or ICR evaluation.  Omit zones ensure that preprinted literals in a zone are not recognized as text.

OMR

Optical Mark Recognition. The process of data selection from a list of options on a document, based on the presence or absence of a mark next to item(s) on that list. See also Mark Sense.

Orientation

The way text is displayed on a page, either vertically (portrait) or horizontally (landscape). The orientation parameters in OCR for AnyDoc allow users to ensure that text in a page reads from left to right as it is being processed, regardless of the text orientation on the page when it was scanned.

Output

Output is the final phase of OCR for AnyDoc processing. Once the data has been captured, validated and verified, both the data and the document images are then delivered to a company’s back-end system.

Output Parameters

Enable the configuration of both ASCII text and images output by OCR for AnyDoc. They can be configured in the form level or the zone level.

Overlay

An image that is superimposed on all data images during verification and/or is archived for a specific master form template.

P

Parameter

A set of tools to help OCR for AnyDoc fine-tune form removal and recognition. It also helps to define rules and output specifications.

Pass 1 Verification

During this phase of OCR for AnyDoc verification, operators view a data image’s questionable characters in the context of the zone and form in which they appear. Pass 1 Verification also allows the operator to correct any recognition rules implemented by rules parameters, mark sense parameters, table link parameters, etc.

Pass 2 Verification

An optional OCR for AnyDoc verification phase that functions either as a method to verify data not examined by Pass 1 Verification, or as a follow-on supplement to Pass 1 Verification.

Patch Code

A parallel pattern of alternating black bars separated by spaces and placed near the leading edge of a paper document. Sometimes used to separate documents and batches or to perform identification.

Permissions

Security measures applied to objects (e.g., database tables, etc.), based on defined user rights.

Pixels

Picture (pix) elements (els). Filled-in dots in a grid that form text or a picture on a computer screen or on printed output.

Process

Process is the hardest working phase in OCR for AnyDoc.  During processing, OCR for AnyDoc separates data from non-data form elements, such as character boxes, lines and background noise. Once the data is separated, OCR for AnyDoc captures the data and validates it against pre-defined business rules.

Q

Quality Assure (QA)

In AnyDocCAPTUREit and OCR for AnyDoc, an additional batch processing phase (off by default) that allows the operator to check and improve the quality of scanned or imported images.
In OCR for AnyDoc, whether and how the quality assurance phase is used depends upon the form family settings.

Questionable Character

A data character with a value undetermined by the recognition engine (where the confidence percent level is below the configured value).

QuickApp™

With QuickApp technology, OCR for AnyDoc users can eliminate key-from-image processes when capturing data from exception or seldom-seen documents – without the need for a template.

R

Reader Response Zone

A type of mark sense zone, Reader Response zones define areas of the form to be evaluated for the presence or absence of a mark, which typically takes the form of a circle around a number.

Registration Zone

The defined area of a document image that allows OCR for AnyDoc to determine the image’s length and width so the program can effectively remove skew from the image and align it to the associated Master Form Template. The Registration Zone consists of two or more registration points on the document, defined by an image, a literal, a cross line and/or data.

Remote Verification

The ability of a human operator to verify, from an off-site location, characters flagged by OCR for AnyDoc as questionable.
Both AnyDocCAPTUREit and OCR for AnyDoc enable access to data verification from a remote location via a LAN or an Internet connection.

Rope and Expand

Roping and expanding magnifies a selected area on a scanned image. The smaller the roped area, the greater the magnification.
During template design, an area must be roped and expanded prior to adding a zone.
During key-from-image processing, roping and expanding text on the document image automatically populates the associated data fields.

S

Sticky Note

A tool within OCR for AnyDoc that verification operators can use, without disrupting verification activities, to notify a supervisor of unexpected results in a particular row or line of data during processing.

String

A sequence of data characters.

Structured Documents

Forms and documents where the desired data are located in static positions on the page across the document type. Examples of structured documents include surveys and vehicle registration forms.
OCR for AnyDoc and AnyDocCAPTUREit specialize in processing structured documents.

Table

A file containing organized data (in rows and columns) on a specific topic, such as Vendor ID Number. As OCR for AnyDoc processes, it can access lookup tables to automatically populate data fields related to the documents and data it captures.

Template

See Master Form Template.

U

Unstructured Documents

Forms and documents where the desired data can be located in varying positions on the page of the same document type. Examples of unstructured documents include invoices and Explanation of Benefits (EOB) forms.
AnyApp technology was developed to process unstructured documents and is found in AnyDocINVOICE and AnyDocEOB.

User Group

A group of OCR for AnyDoc users with the same access rights. The OCR for AnyDoc administrator defines the user groups and grants rights to them.

V

Verify Phase

The phase of OCR for AnyDoc processing where data characters flagged as questionable by OCR for AnyDoc get verified, either by a separate recognition engine or by a human operator. This is done during the Verify phase of OCR for AnyDoc processing.

W

Work Flow Manager

The control panel for all production-level batch processing performed by OCR for AnyDoc.

Z

Zone

An area in the Master Form Template defined as the location of a specific data type. Each zone type is designated by a separate zone boundary color.