Locator anchor zone recognition

Using a locator as input for an anchor enables you to use a field on the form as a reference point for other extraction fields.

When you configure your project, the location of the reference locator must be above the Advanced Zone Locator with the anchor, in the Project Tree structure. This is because the order of locator methods determines the order of extraction, and the result of the reference locator is needed in order to locate the associated anchor.

You may have a document that does not have a suitable structural, graphical, or text anchor. You do however, have a piece of information that differs on each document, but is always present. The Locator anchor enables you to use that field as an anchor to improve recognition of your documents.

The input locator method type must be locator that returns a single alternative on the reference image. This means that locators that return multiple subfields such as the Table Locator, Line Item Matching Locator, Advanced Zone Locator, and the group locators, to name a few, are not supported as anchor input.

For example you can use a Format Locator, a Barcode Locator, a Script Locator, and any other locator that returns a single piece of data.

For example, you have a group of documents that have a bar code at the beginning of the customer data on the form. The data above the bar code is not of interest for extraction, so you can use the bar code as an anchor point for the relevant information that you are extracting. Set up a Barcode Locator and then use that as input for the Locator anchor.

Similarly, a document always contains a 10-digit customer identification number, that differs on every document. The format of the customer number never changes, and its location is consistent. You can use a Format Locator to find this customer number and then use the Format Locator as input for a Locator anchor. Doing so can improve the overall recognition and extraction results for your documents.