The typical data capture process involves the following four steps:
1. Loading images – At this step, document images are added to the project
2. Recognition – At this step, the data on the images are recognized
3. Verification – At this step, the recognized data are verified
4. Export – At this step, the verified results are saved
Each step has its own button on the toolbar:
Loading images
The arrow to the rights selects how images will
be obtained. You can Load Images, Scan Images, or Import Images using one of the
available image import profiles. The caption of
the button changes to reflect your choice.
Recognition
Starts the recognition process. Click the arrow to
the right to select the Analyze or Match
Template command.
Verification
Starts the verification process. Click the arrow to
the right to select the Re-check Rules command.
Export
Export the data based on the template settings.
Click the arrow to the right to select the Export to File or Export to Database command.
Loading images
First, you must select a project:
1. Click
2. Select a batch or create a new batch into which the images are to be loaded. To create a new
batch, right-click anywhere in the main window and select New Batch on the shortcut menu.
If you try adding images to a project that contains no batches, a new batch will be created
automatically.
Now you must load the images into the batch. There are several ways of loading images:
1.You can load existing image files. To load existing images, click the arrow next to the Import
Images… button and select Load Images... Alternatively, you can press the Ctrl+O keyboard
shortcut.
2. You can scan paper documents. To scan paper documents, click the arrow next to the Import
button and select Scan Images... The program will prompt you to select a scanner.
3. You can import images using one of the image import profiles that were earlier created by the
administrator.
If there are image import profiles available, their names will be displayed on the drop-down
menu of the Import button. Select the desired profile to start importing images.
Clicking Import Images… opens the Select Import Profile dialog box. Select the desired
import profile and click Import to start importing images.
Once you select an image import profile, its name will be displayed on the Import Images…
button so that you don’t have to select this profile from the list next time you want to use it.
Images may be imported in background mode if an appropriate import profile was created. Selecting
this type of profile automatically loads images into the batch from a dedicated “hot” folder.
After you add the images, unprocessed pages appear in the list.
Recognition
Clicking the Recognize. Button starts the recognition process.
The recognition process may be launched automatically as soon as images are added into the batch.
To enable this option, select Tools>Options, click the Document Processing tab, and check
Recognize added images automatically.
The Confidence Level column displays the percentage of reliably recognized characters.
Once the recognition process is complete, you may verify the results.
Verification
Verification consists in checking the recognized data for errors.
For multi-page documents, you must first check if the pages have been correctly assembled into
documents. Then the recognized data are verified using group and context verification modes. You
can also verify the data in the document window. Rules are also checked at this stage.
1. Checking document assembly. This check is not required when processing one-page documents —
you can go directly to data verification.
For multi-page documents, you must first check if the program has assembled the pages into
documents correctly.
If the order of pages in a document does not match the order specified for this document or if the
values of the key field are not identical on all the pages, the document is marked with a red flag and
an error message is displayed in the document window.
If this is the case, first make sure that pages were not mixed up at
the scanning stage. Many assembly errors can be corrected by
simply changing the order of pages.
You can check document assembly in thumbnail view (Figure 1). In this view, you can change the
position of pages and even move them between documents with the mouse.
If key field values are used to ensure correct document assembly, the values of the key fields are
displayed below the image of each page (Figure 1). If key field values are not identical on the pages
of the same document, they will be displayed in red.
A mismatch of key field values may occur if they have been recognized or filled out incorrectly.
Verify the key field values. If the mismatch persists, the mismatching pages are from different
documents. Find the pages with identical key field values and assemble them into documents.
Note: To change the scale of the thumbnail images, hold down the CTRL key and scroll the mouse
wheel.
Figure 1. Document pages in thumbnail view
Next, start the verification process by clicking Run Verification. First, the group verification
window opens. Once the group verification is finished, the context verification window opens.
2. Group verification groups together the images of characters that have been recognized identically.
Identical characters (e.g. digits 1 in Figure 2) are shown in groups so that you can confirm the
obviously correctly recognized characters and postpone the incorrect or dubious ones until the next
verification stage.
If you have doubts about a character:
1. Right-click the character and select Show Character Image on the shortcut menu or press
F2. The image of the field containing the character being verified will be displayed.
2. In the verification window, select View > Field Image > Show Field Image or press the
Ctrl+I keyboard shortcut. The verification window will be split into two panes and the
appropriate field will be displayed in the bottom pane when you rest the mouse cursor on a
dubious character.