Text matching

You can add flexibility to your applications by using text matching to identify pages and locate data.

You used fingerprints to identify pages and recognition zones to locate data on those pages. The one exception was when you used the RegExFind action to locate the Total Cost field when you process line item grids. You saw in that example how you can use text matching to locate data that is not in a predictable location on the page.

To use text matching, you first do full page OCR on the incoming page. You can then search the recognition results for specific text. For example, if a page contains the words Car rental agreement, then the chances are high that it is a car rental agreement page. Similarly, if you locate a string Pickup date, then the chances are high that beside or below the string is the actual pickup date.

You can identify pages and locate data by using text matching. You can update the TravelDocs application to identify and process a new car rental agreement page by using text matching.