| Oracle® Distributed Document Capture/Oracle® Document Capture Administrator's Guide Release 10gR3 E13871-01 |
|
 Previous |
 Next |
Home > Administrator's Guide > Oracle Document Capture Ove... > About Index
| Oracle® Distributed Document Capture/Oracle® Document Capture Administrator's Guide Release 10gR3 E13871-01 |
|
Previous |
Next |
Home > Administrator's Guide > Oracle Document Capture Ove... > About Index
This section covers the following topics:
Users index images by assigning one or more values to them in index fields. During commit, Index takes all images with the same index values and merges them into a document. When it encounters a different value for any of these fields during the commit process, it creates a new document. After images are committed, they are removed from the batch and no longer available.
The following indexing guidelines apply.
A document is created from more than one image when all of the index field values among the images are exactly the same. If one image has three fields filled in and another image has two fields filled in with the same values, two documents are created.
An image is indexed if there is a value in any of its index fields.
Unindexed images (those without any fields filled in) remain in the batch when users commit it.
Required fields must be filled in before their images can be committed. Required fields have an asterisk (*) in front of their name in the indexing screen to users.
An index profile contains all the indexing settings you want used when users index particular batches. Through its settings, you:
identify the file cabinet from which scanned batches are indexed and into which they are stored,
identify fields to be indexed and indexing methods to be used,
can map an external database from which users can look up index values,
can identify how separator sheets are used,
can set up certain fields such as a scan date to be autopopulated,
can set up image highlighting using the Zone Editor.
Unlike scan profiles, index profiles are typically stored in a central location on the network. This allows many indexing users to use the same index profile.
This section provide general tips for optimizing indexing.
When scanning, separate your single and multiple page documents into different batch types and create two index profiles that can be optimized for each document type. You can then use separator sheets in the multiple page documents to index them more quickly.
The Sticky Fields feature is useful in batches where you have a mixture of single page and multiple page documents. When a field is sticky, the index field value a user enters is automatically copied to the same field on the next page when paging through the batch. (Sticky fields also work when paging backwards in a batch.) If you set all fields as sticky, index values are automatically copied onto the new page; if the user determines that the new page does not belong to the same document, he or she simply enters new values and that page becomes a new document. For details about setting sticky fields and specifying when a field is sticky, see "Sticky Index Fields".
The built-in database lookup feature is very powerful as it can eliminate data entry of certain fields and ensure data integrity. For example, in an accounts payable application, you could data enter the purchase order number and perform a database lookup to retrieve the vendor number, vendor name and PO date from your accounting system.
Use pick-lists whenever you want to standardize specific field values. For example, you might create a pick-list of ten document types and require that a document be one of those ten types by locking the field.
Order your indexing fields to coincide with the way the indexing user will read the image. This allows the indexing user to process an image naturally.
When the Scroll Lock key is present and active on users' keyboards, each time they press the Tab key, the index fields shift up and the topmost field is hidden from view. (Pressing Shift-Tab shifts index fields down.)
If you are processing images that are form-based, take advantage of the auto-highlighting feature by defining field zones in the Zone Editor. This forces Index to automatically highlight an area on the image, allowing the user to see and process the field quicker. Use this feature in conjunction with the SCROLL LOCK key to reduce end-user eye movement. Reduction in eye movement can help to minimize indexing user stress and fatigue.
If needed, use the input format feature to ensure that data users enter is formatted consistently. When the user moves to another index field, Capture changes the data if needed to match a selected format.
You can apply special field properties to index fields. For example, applying a pick-list to an index field provides a standard list from which users select. Special field properties include:
A pick-list provides users with two or more values from which to choose when indexing documents. You create pick-lists in Capture Administration, then select them in the index profile.
In a parent/child relationship, a parent field is linked to two or more child pick-lists, one of which is displayed after the user makes a selection in the parent field. For example, a parent pick-list might list departments and its child pick-list document types; if the user chooses Sales from the parent pick-list, the child pick-list might list document types specific to Sales. You create pick-lists and their relationship profiles in Capture Administration, then select the relationship profile and pick-lists in the index profile.
A mask is a filter that controls what users can enter into the index field. For example, a mask might allow users to type numbers but not letters. If the user types a character that does not conform to the mask, Index does not accept it.
The Input Format option provides a way to ensure that data entered by users matches a certain format. When the user moves to another index field, Capture changes the data, if needed, to match the specified input format.
You can configure a scan profile to detect bar codes when images are scanned, and then an index profile to automatically apply the detected value to an index field. If the page contains more than one bar code, Index displays their values in a pick-list to the indexing user.
As an alternative to indexing, images can also be indexed using Recognition Server. This component automates indexing through bar code recognition technology.
Use the Auto Populate tab to set up automatically indexed fields. You can configure Index to automatically fill (populate) selected index fields with certain values. For example, you might want to automatically populate a date field with the batch's scan date instead of having users type it in. Or, you might want to provide a default value for a field that the user could change if needed.
When an index field is sticky, an indexing user specifies a value in an indexing field, moves to the next page and the value is automatically pasted into that indexing field (if blank) on each subsequent page the user displays.
Sticky fields can significantly speed up indexing. In a common scenario, all fields are set as sticky; a user enters values on the document's first page, then quickly views each remaining document page as values are automatically applied.
The sticky field on the destination page must be blank.
Sticky fields are sticky only when users navigate using the Next Page or Previous Page toolbar buttons, and not when using thumbnails or typing page numbers in the Goto Page field.
To provide flexibility, a When option controls when the field is sticky. The choices are ordered from most restrictive to least:
All index fields blank (the default; most restrictive): The field is sticky only when none of the destination page's fields contain values.
Page partially indexed: The field is sticky only when the page is not considered fully indexed. For example, one or more index fields may contain values, but a required field may be missing.
Always (least restrictive): The field is sticky regardless of the page's other index field values.
How Autopopulated Fields Affect Sticky Fields
Because autopopulated fields can affect whether a page is considered indexed and therefore whether a field is sticky, an option titled Allow autopopulate field(s) to index page allows you to control this.
When NOT selected (default), autopopulated fields are considered blank.
When selected, autopopulated fields are considered indexed fields.
Note that, when the user moves to a destination page, sticky field logic is applied BEFORE values are autopopulated. So an autopopulated field affects the sticky field only if the autopopulated field previously had a value.
When All index fields blank is chosen in the Sticky When field, the Allow auto populated... field affects whether all fields are considered blank and consequently, whether the sticky field is sticky. When deselected, all fields are considered blank, so the field is sticky. When selected, the autopopulated field is considered indexed, all index fields are not blank, so the field is NOT sticky.
A database lookup searches an external database for a value the user specifies and populates index fields from a matching database record. For example, users might perform a database lookup after entering an account number, and the remaining index fields would be populated from the matching database record.You create a database lookup in Capture Administration, then select it in the index profile. For more information about database lookups, see "About Database Lookups".
You can use separator sheets to determine where documents begin and end in a batch. Separator sheets speed up indexing: instead of viewing each page to determine if it belongs with the current document or is part of a new document, the user enters index values once for each field, applies those values to all pages in the document, and then quickly jumps to the start of the next document.
To index by separator sheet, batches must have been scanned using patch code or bar code detection in Scan.
When separator sheet indexing is enabled and the user selects the File to Separator command from the Batch menu, Index copies all index values from the current page to all pages between the separator sheets, then advances to the next document (separator sheet).
Index's zone feature allows you to highlight certain areas (zones) of an image with a turquoise-colored box to guide the user when indexing. Highlighting fields to be indexed in this way can increase user efficiency by reducing eyestrain and making fields easier to read. The zone feature is particularly useful when users are indexing forms.
You set up zones for an index profile using Index's Zone Editor. When indexing, users can click the Display Toggle Zone button on their toolbar to turn zones on or off.
To set up zones, you need an image file to use as a template. It is important that the resolution of this image is the same as the resolution that will be used at scan time.
You can configure Index to capture text in a zone and automatically enter it in a selected index field. If OCR is active for an index profile, users can also perform OCR on-the-fly: they right-click and drag an area on the image, and Index will capture its text and insert it in the selected field.
Important points about the OCR feature:
You activate OCR for a zoned field by setting up the field's zone and turning on OCR by selecting the OCR button on the Zone Editor toolbar.
OCR works well in conjunction with a macro that parses the text captured. This allows you to define a zone with padding around the text to capture, and then extract only the characters you want. For example, you might do this when capturing Social Security numbers, capturing a larger area, and then using a macro to filter out all characters except the Social Security number.
If OCR is enabled for any field, users can perform OCR on-the-fly, even if that field is not enabled via a zone, and even if zone display is turned off. Users do this by rubber-banding an area on the image while holding down the right mouse button.
Index performs zonal OCR on an image only if it is not indexed (does not contain values in its index fields). If you want Index to perform OCR on an indexed image (on all fields for which OCR has been defined), choose the OCR Zone(s) command from the Edit menu, or press F5.
It is important that users verify OCR entries, since it can result in inaccurate values.
